NATIONAL INSTITUTE OF MENTAL HEALTH ba i ~ NEUROSCIENCE OF MENTAL #8 ie ] é - — ara VASA a ie ea) os { Ps SY ae U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Alcohol, Drug Abuse, and Mental Health Administration ''NIMH Neurosciences Research Branch ADVISORY PANEL MEMBERS George K. Aghajanian, M.D. Yale University School of Medicine Julius Axelrod, Ph.D. National Institute of Mental Health Jack David Barchas, M.D. Stanford University School of Medicine Erminio Costa, M.D. National Institute of Mental Health Peter B. Dews, Ph.D. Harvard Medical School Arnold J. Friedhoff, M.D. New York University School of Medicine Patricia Goldman-Rakic, Ph.D. Yale University School of Medicine Roger A. Gorski, Ph.D. University of California School of Medicine, Los Angeles John A. Harvey, Ph.D. University of Iowa Walter F. Heiligenberg, Ph.D. University of California, San Diego Steven A. Hillyard, Ph.D. University of California, San Diego Barry J. Hoffer, M.D., Ph.D. University of Colorado Medical School Leo Hollister, M.D. Veterans Administration Medical Center, Palo Alto Susan Iversen, Ph.D. Merck Sharp & Dohme, Ltd. Eric R. Kandel, M.D. Columbia University College of Physicians and Surgeons David J. Kupfer, M.D. University of Pittsburgh School of Medicine Bruce S. McEwen, Ph.D. Rockefeller University Allan F. Mirsky, Ph.D. National Institute of Mental Health Donald W. Pfaff, Ph.D. Rockefeller University Donald J. Reis, M.D. Cornell University Medical Center Larry W. Swanson, Ph.D. Salk Institute EVALUATION PANEL MEMBERS Ralph N. Adams, Ph.D. University of Kansas Albert J. Aguayo, M.D. McGill University Raymond T. Bartus, Ph.D. American Cyanamid Company Henri Begleiter, Ph.D. State University of New York, Downstate Medical Center Michael J. Brownstein, M.D., Ph.D. National Institute of Mental Health Verne S. Caviness, M.D., Ph.D. Massachusetts General Hospital Victoria L. Chan-Palay, Ph.D. Harvard Medical School David H. Cohen, Ph.D. State University of New York, Stony Brook Joseph T. Coyle, M.D. Johns Hopkins University School of Medicine Samuel A. Deadwyler, Ph.D. Bowman Gray Medical School Anke A. Ehrhardt, Ph.D. New York State Psychiatric Institute Everett H. Ellinwood, Jr., M.D. Duke University Medical Center Salvatore J. Enna, Ph.D. University of Texas Medical School, Houston John N. Fain, Ph.D. Brown University John D. Fernstrom, Ph.D. University of Pittsburgh School of Medicine William J. Freed, Ph.D. National Institute of Mental Health Michael S. Gazzaniga, Ph.D. Cornell University Medical Center J. Christian Gillin, M.D. University of California, San Diego William T. Greenough, Ph.D. University of Illinois, Champaign Philip Groves, Ph.D. University of California, San Diego Peter J. Hand, V.M.D., Ph.D. University of Pennsylvania School of Veterinary Medicine Boyd K. Hartman, M.D. Washington University School of Medicine James N. Hayward, M.D. University of North Carolina Medical School Stephen F. Heinemann, Ph.D. Salk Institute Edward Herbert, Ph.D. University of Oregon Myron A. Hofer, M.D. Albert Einstein College of Medicine David Housman, Ph.D. Massachusetts Institute of Technology Paul Insel, M.D. University of California, San Diego Barry L. Jacobs, Ph.D. Princeton University Lily Yeh Jan, Ph.D. University of California, San Francisco Thomas M. Jessell, Ph.D. Harvard Medical School Arthur Karlin, Ph.D. Columbia University College of Physicians and Surgeons Harvey J. Karten, M.D. State University of New York, Stony Brook Darcey B. Kelley, Ph.D. Columbia University David C. Klein, Ph.D. National Institute of Child Health and Human Development George F. Koob, Ph.D. Scripps Clinic and Research Foundation Daniel E. Koshland, Jr., Ph.D. University of California, Berkeley Edward A. Kravitz, Ph.D. Harvard Medical School Dorothy T. Krieger, M.D. Mt. Sinai Medical Center Irving Kupfermann, Ph.D. Columbia University College of Physicians and Surgeons Philip W. Landfield, Ph.D. Bowman Gray School of Medicine Victor G. Laties, Ph.D. University of Rochester Medical Center Lee E. Limbird, Ph.D. Vanderbilt University Richard E. Mains, Ph.D. Johns Hopkins University School of Medicine Arnold J. Mandell, M.D. University of California, San Diego John F. Marshall, Ph.D. University of California, Irvine R. Bruce Masterton, Ph.D. Florida State University Jeffrey F. McKelvey, Ph.D. State University of New York, Stony Brook Herbert Y. Meltzer, M.D. University of Chicago School of Medicine Michael Menaker, Ph.D. University of Oregon Klaus A. Miczek, Ph.D. Tufts University Perry B. Molinoff, M.D. University of Pennsylvania School of Medicine - continued inside back cover '' THE NEUROSCIENCE —_OF MENTAL HEALTH A Report on Neuroscience Research Status and Potential for Mental Health and Mental Illness . From Panels of Scientists Representative of Contributing Disciplines U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Alcohol, Drug Abuse, and Mental Health Administration ,_ National Institute of Mental Health ' 5600 Fishers Lane Rockville, Maryland 20857 ''#4 204355L i “Pain of mind is worse than pain of body." Latin Proverb All material appearing in this volume is in the public domain and may be reproduced or copied without permission from the Institute. Citation of the source is appreciated. Library of Congress Catalog Number 84-601129 ADM Publication No. 84-1363 Printed 1984 ''FOREWORD The mission of the National Institute of Mental Health (NIMH) is to support research on the diagnosis, etiology, treatment, and ultimate pre-— vention of mental illness. In 1981, the NIMH reorganized its research divisions and created within the Division of Extramural Research Pro-— grams (DERP) a new and separate Neurosciences Research Branch. These organizational changes reflect a renewed emphasis on the research mis— sion of the NIMH, as well as a recognition of the centrality of basic brain sciences within that research mission. The Neurosciences Research Branch has the responsibility for supporting basic research to understand normal brain function and the neural substrates of behavior, especially mental disorders and maladaptive behaviors. In all areas of intellectual endeavor there is a periodic need for a critical assessment that can serve to guide investigators, science admin-— istrators, and funding agencies. Given the rapidity of developments in neuroscience research and the formation of a Neurosciences Research Branch at NIMH, the time was propitious for "the taking of an inventory" on neuroscience research issues. This report is the result of a series of evaluation meetings convened by the Neurosciences Research Branch with a perspective toward delineating high-priority research areas for the 1980s. The information contained herein should serve as a stimulus to: (a) clinical researchers, (b) neuro— scientists, (c) other scientists whose attention to mental health research issues would significantly advance the knowledge base in the mental health field, and (d) young investigators contemplating research careers. For the science administrator, this report will serve as a guide for setting goals and as a yardstick by which to measure accomplishments. For the public, this report provides some insight into basic science research ef- forts geared toward solving the problems of mental illness. Through the publication of this volume, the NIMH is establishing a major commitment to neuroscience research and hopes to gain increased interest and enhanced activity in the conduct of neuroscience research in the context of mental health and mental illness. Larry B. Silver, M.D. Acting Director, National Institute of Mental Health iii '' ''PREFACE The formation of a Neurosciences Research Branch within the National Institute of Mental Health, chargéd with a specific mandate for the sup— port and development of neuroscience research, provided a unique op-— portunity for defining and conceptualizing those areas within neuroscience research that have the greatest potential of contributing to mental health research. The initial tasks in this process were to define the state of neuroscience research in relation to mental health issues and then to delineate further research directions for support by the NIMH. In order to accomplish this, the Neurosciences Research Branch sought the expertise and guidance of distinguished scientists in the several disciplines that collectively comprise the field of neuroscience. These scientists, whose names appear on the inside covers of this volume, participated in a series of small workgroups. The outcome of their thoughtful deliberations is presented in this monograph in a condensed, restructured format. The evaluation process was designed to examine critically the basic assumptions, methods, and goals of neuroscience research as related to issues in mental health and mental illness. The field of neuroscience re— search is extremely broad. We, therefore, did not attempt in this current effort to cover completely all research areas. This should not necessarily be interpreted as a lack of interest in these areas or as a statement of their relative importance. Emphasis was placed on the latest research findings by selecting specific areas to illustrate areas of greatest poten— tial and weaknesses. The evaluation process and this report focus on critical issues and areas that must be dealt with if neuroscience research is to be effectively exploited in both clinical and nonclinical applications. The report also defines those areas most critical to advancing the theo- retical understanding of brain—behavior relationships. The format of the evaluation process—-—mostly individual assessment and discussion--was not one suited to the citation of previous published papers. Thus references are not a part of this work. In addition, the dy- namic and heuristic quality of the discussions that characterized the workshop interactions could not be captured in a volume such as this. It is hoped, however, that the multi- and interdisciplinary nature of the dis- cussions will be reflected in the text that follows. The issues discussed and the recommendations contained in these chapters are specific to the topic considered in each chapter. Several major themes, however, emerged from all the scientific panels, regardless of the scientific area under discussion. These common themes are briefly reviewed here to provide a context for the substantive issues contained in the text. Neuroscience research is currently on the brink of discovery of the underlying neuronal functions that are key to understanding the brain mechanisms serving behavioral expression. This progress has been made possible because of the inter- and multidisciplinary approaches inherent in neuroscience research and because of the technological breakthroughs made in several other fields, e.g., molecular biology, physics, computer sciences, etc. It is necessary to caution, however, that the “rush to go molecular" that characterizes much of this work must be tempered by a keen awareness of behavior. The ultimate goal, after all, is to understand how the brain orchestrates its complex systems of neurons and molecular machinery in the production of behavior-—both normal and aberrant. In working toward the goal of understanding brain function in relation to human behavior, studies using higher vertebrates are, of course, the most relevant. Because of the complexity of the nervous system the use of lower vertebrates and invertebrates is an extremely important and Vv ''valuable approach. The conservation of biological mechanisms across species has been amply demonstrated, and the information gained from studies with simpler systems is valid and informative. Cellular and syn- aptic information from simple animals helps our conceptualizations to be concrete and suggest principles for application in more complex nervous systems. For example, classical studies of neurotransmission using frog neuromuscular junction or Torpedo electric tissue illustrate the impor- tance of selecting an appropriate and accessible experimental prepara- tion. Restraint, however, must be exerted against extremes in taking a reductionist approach to synaptic events and brain function. While there are a multitude of valid reasons to study invertebrates and lower verte- brates, it is also necessary to keep in mind that many higher brain func- tions are absent in phylogenetically primitive animals. For example, there are probably unique mechanisms for long-term memory that appear only late in phylogeny. Indeed, recent work suggests that the hippocampus, a brain structure that first appears in lower vertebrates, confers upon organisms a totally new kind of information storage capacity. To under— stand normal and abnormal behavior, research needs to be pursued at all levels of analysis. The major determining factor in choosing the prepara— tion for study should be the ability of the species chosen to test most directly the hypothesis under consideration. Although our ultimate research goal is knowledge about the neural basis of psychopathology, it is often impossible to draw a definitive line be- tween normal and abnormal behavior. It should be acknowledged, there- fore, that information gained about the organizational control of normal behavior will increase our understanding of the abnormal and, most im- portant, is a crucial first step in defining abnormal behavior. Thus, psychopathology should not and cannot be approached as an isolated phe- nomenon but rather must be studied in the context of normal behavior. The majority of experimental approaches available for analyzing brain function cannot be used in humans; therefore, animal models of psycho— pathology are invaluable for neuroscience research. Fortunately, humans have neuronal systems and behavioral repertoires common to many other species, so that knowledge gained about the neural processes involved in behaviors of other species is applicable to major areas of normal human behavior and psychopathology. These models are not meant to be replicas of psychopathology or to encompass an entire syndrome, but only to em- body in an experimentally approachable form certain features or symptoms of the human condition. Finally, psychopharmacological agents are of particular importance to basic neuroscience research. They are valuable tools for the experimental manipulation of behavior and neurochemical systems. Discovery of their mechanisms of action at the molecular and supramolecular levels can also enhance our understanding of those neuronal mechanisms that are amen-— able to therapeutic manipulation in psychiatric disease states. This understanding, in turn, increases the potential for developing psycho- pharmacological agents with more specific sites of action. Such changes will provide improved treatments for the mentally ill. This monograph is the product of a collective activity involving the conscientious efforts of many people dedicated to the support and conduct of neuroscience research. Grateful acknowledgment of their contributions is made to members of the scientific panels and to members of the NIMH Neurosciences Research Branch, DERP, named on the inside cover of this volume. The assistance of the Office of Policy Development, Planning, and Evaluation is acknowledged. Appreciation and gratitude is expressed to Dolly Gattozzi, who unrelentingly pursued the overwhelming task of integrating the individual and group reports from the scientific panels. Stephen H. Koslow, Ph.D. Chief, Neurosciences Research Branch National Institute of Mental Health ''CONTENTS FOREWORD PREFACE INTRODUCTION CHAPTER I. NEURAL DEVELOPMENT AND FUNCTIONAL ORGANIZATION Neurogenesis and Development Phenotype Control Cell-cell Recognition Trophic Factors Neuronal Death Hormonal Requirements Neural Transplants and Plasticity Aging Neural Circuitry Functional Mapping Biochemical Analysis Receptor Localization Modification of Brain Development and Function Receptive Mechanisms Hormonal Influences Circadian Circuitry Nutrition and Diet Experience Recommendations for the Future CHAPTER II. NEUROTRANSMISSION: BIOLOGICAL AND PHARMACOLOGICAL ASPECTS Transmitter Identity Acetylcholine Amino Acids Amines Peptides Transmitter Synthesis Transmitter Storage and Release Transmitter Coexistence Transmitter Inactivation Transmitter Receptors Acetylcholine Receptor Receptor Function and Type Receptor Pharmacology Postreceptor Coupling Mechanisms Ion Channels Genetics of Neurotransmission Molecular Approaches Developmental Studies Structural Determination Neuropeptides Drug Development Mental Illness Recommendations for the Future page iii COMWAAAANUNAA LH ''CHAPTER III. HORMONES AND NEUROPEPTIDES IN THE CONTROL OF BEHAVIOR Hormonal Effects Invertebrates Genomic Regulation Site of Action Modulation of Neuronal Function Development Homeostatic Processes Feeding Behavior Drinking Behavior Temperature Regulation Reproductive Behaviors Sleep Circadian Modulations Recommendations for the Future CHAPTER IV. BEHAVIORAL ANALYSIS OF PSYCHOACTIVE DRUGS Behavioral Pharmacology Ethopharmacology Electrophysiological Analysis Neurobehavioral Pharmacology Biogenic Amines Neuropeptides Drug-—Induced Models of Psychopathology Geriatric Psychopharmacology Behavioral Toxicology Recommendations for the Future CHAPTER V. THE NEURAL BASIS OF PSYCHOPATHOLOGY Recombinant DNA and Family Mental Illness Genetic Linkage Imaging Techniques Evoked Brain Potentials Brain Function in Psychopathology Neurotoxins Neural Transplants Developmental Psychobiology Developmental Stage Mother-Infant Interaction Genetic Approaches Learning and Motivation Sexual Differentiation Higher Order Integrative Processes Information Processing Learning Models of Psychopathology Memory Language Neuropathology Pharmacological Models Social Models Recommendations for the Future ABBREVIATIONS viii ''INTRODUCTION This report is the result of the efforts of the Neurosciences Research Branch staff, who planned and implemented a multifold discussion and report process designed to forecast and recommend neuroscience—mental health research areas offering the greatest current potential. Using the Institute's existing portfolio in basic biological research as a point of departure, the three research sections (Biobehavioral, Neuro- biological, and Psychopharmacological) within the Neurosciences Research Branch individually undertook evaluations focusing on the range of research interests under their purview. The final reports were then integrated and edited by pro- gram staff with the aid of a science writer to produce this document. The Evaluation Process For each of the sections of the Neurosciences Research Branch, the evaluative process was accomplished in three stages. The first stage involved convening an advisory panel of senior scientists, representing the full scope of the program's interests but selected primarily on the basis of their broad research perspective and general knowledge of the NIMH mission. The task of the advisory panel was to review briefly the grant holdings of the section, identify sub- areas within the program that would be amena— ble to in-depth analysis, and nominate scientists who would be appropriate to evaluate each of the subareas identified. The evaluation panel nominations were made in the context of the latest developments in each area, in an effort to ensure that emphasis would be placed on future directions and areas of greatest promise. The second stage of the evaluation process consisted of the evaluation panel meetings. Each panel was chaired by a member of the advisory panel to maintain continuity. In advance of the evaluation panel meetings, each panel member prepared: a review of the state of the science in his or her research area, an evaluation of pro- gram holdings in the context of this review, projections of the future directions of research, a statement about special needs of the field, and specific recommendations to program staff about how best to foster research in the area. These materials were individually presented and discussed in detail at each panel meeting and then a report was prepared by each evaluation panel, representing the views of all of the panel members. Evaluation panel reports were edited by staff and recirculated for corrections and comments before being prepared in their final form for the review and integration by the advisory panel. The third and final stage of the evaluative process involved reconvening the members of the original advisory panel to assimilate and inte— grate the reports of the evaluation panels and to prepare a final report. This final report stressed not only the needs raised by the individual panels but also included needs of the field as a whole that could be served by an integrated program effort. These overarching needs of the field are presented below. Program Support In order to take advantage of the many new opportunities for advancing knowledge about the biological bases of mental health and illness, panelists recommended a number of ways in which NIMH could help meet current needs of the research community. These may be summa-— rized as follows. Innovation in research should be accorded a high priority. (1) Grant applications require ap— proval for periods of time sufficient to allow for ventures into truly new and untried areas. Ap— proval periods of only 2-3 years encourage in-— vestigators to propose work that is already ongoing. (2) Existing research projects could be extended to include short-term pilot or feasi- bility studies that are targeted exclusively to-— ward innovative aspects of the research. Studies to develop new technologies, to transfer tech- nologies from other disciplines, and to test new concepts are particularly encouraged. (3) Grant applications could be modified to include a special section on “innovativeness," thus high— lighting, for the reviewers and for science ad— ministrators, the importance of this aspect of the proposed research. More than ever before, it is necessary to foster interdisciplinary research. Breakthroughs in various scientific disciplines must be brought to bear on mental health problems if significant progress is to be made. For example, techno-— logical advances in molecular biology hold great promise and must be applied to issues in mental health research. (1) The degree of heritability of certain forms of mental illness is clearly quite significant. Thus, the application of genetic ''linkage methods to mental illness is essential during the next decade. The most important priorities in this area at present are to develop a library of markers for the human genome, while simultaneously conducting field studies to iden— tify family groups in which linkage studies can subsequently be carried out. Kindreds of subjects who have disorders of psychiatric interest should be established and utilized as molecular genetic techniques are applied to mental health re-— search. (2) In vivo imaging techniques should be used to improve and validate, in humans, in- formation that has been well established in other species. Increased collaboration should be sup- ported between highly qualified neurobiologists, or psychobiologists, and research centers uti- lizing in vivo imaging techniques. In particular, efforts should be made to improve imaging technology so as to eliminate or control for artifacts or adventitious effects. Established investigators as well as young re— searchers require special training experiences to master new approaches to problems and to en— able collaboration with researchers in allied disciplines. Although individual scientists cannot be expected to master all of the disciplines of neuroscience equally well, the joint application of approaches is necessary. This requires that pre-and postdoctoral training in the neuro- sciences be multidisciplinary. It is recommended that in the case of postdoctoral fellowships the award be granted for 3 years to allow fellows to gain adequate exposure and experience in several research approaches. There is also a need to provide opportunities for established researchers to acquire the skills needed to take advantage of new technologies and approaches. Mechanisms for short-term training experiences with rela— tively moderate dollar amounts should be de- veloped by NIMH to afford such investigators in the field of mental health this opportunity. Collaborative efforts between basic scientists and _ clinicians create a setting in which the cross—fertilization of ideas will provide new in- sights into the pathophysiology of mental ill- nesses. Training experiences should thus be pro- vided to ensure a full cadre of investigators who have both basic science and clinical competence. In this regard, academic institutions should be encouraged to provide training programs mod- eled after those provided by the clinical as- sociates program in the Intramural Research Program of NIMH and the Mental Health Clinical Research Centers. Another mechanism for fos- tering communication between basic and clinical researchers is the workshop. Timely topics of particular relevance to basic scientists and cli- nicians at the present time are suggested for workshops. Access to specialized technological resources should be ensured to investigators in a wide variety of research settings. For example, NIMH should support appropriate mechanisms and re- lationships in order to facilitate access to mi-— crosequencing facilities, cDNA probes for hy-— bridization studies, and antibodies for classical neuropeptides. Contracts, small grants, and special relationships between NIMH and inves— tigators who have access to these specialized resources should all be considered. Additionally, funding should be made available for high— technology instruments, e.g., brain scanners of various types, for shared use at large, high-— quality research centers. It is further suggested that a mechanism be established for the creation and shared use of large breeding colonies for animals that offer special advantages for re— search in neuroscience. Neuroscience Research Centers. Current de-— velopments within the neurosciences and the basic sciences generally make increasingly evi- dent: (a) the promise held by the newer tech- nologies of neuroscience and allied disciplines for the solution of research problems relevant to mental health and (b) the array of specialized fields required to clarify the relationships between cellular brain processes and _ the complex behaviors of the intact organism. For the field of mental health then, the scientific opportunities offered by integrated, multi- disciplinary research appear to hold great promise. It is recommended that the NIMH establish a mechanism for funding Neuroscience Research Centers which would create scientific opportunities for the promotion and integration of approaches with the cross-fertilization of scientific disciplines. Information and substance-distribution clear— inghouses should be established to aid investi- gators in locating and obtaining new reagents such as peptides and monoclonal antibodies as well as special strains of animals and micro- organisms. The number of neurobiologically im-— portant substances deriving from neuroscience research is increasing rapidly, and the expense and logistics of providing these materials to le— gitimate scientific laboratories are often pro— hibitive. The availability of support from NIMH specifically for the production and distribution of such substances could be a great benefit to research. Similarly, much effort is involved in producing particular strains of mutants and ho- mozygous organisms, and some mechanism to increase the availability of these animals would be of considerable benefit to biological research generally. Brain banks should be supported through a mechanism that can provide adequate numbers of specimens from well—diagnosed subjects for use by investigators. One possible mechanism would be to support brain banks attached to Mental Health Clinical Research Centers that are already involved in careful diagnosis of patients with various psychiatric disorders. Such a brain bank should also include a tracking sys-— tem for maintaining contact with families of ''well—diagnosed subjects prior to the death of the subject. The Report The chapters that follow represent summaries of the panels' deliberations and recommenda— tions on: neural development and functional organization, biological and pharmacological as- pects of neurotransmission, hormonal/neuropep-— tide control of behavior, behavioral analysis of psychoactive drugs, and neural basis of psycho-— pathology. In each chapter, state—of-the-science summaries are integrated with the panels' pro- jections of future directions and recom-— mendations for program priorities. At the end of each chapter, some of the most important of these recommendations are highlighted. ''CHAPTER I NEURAL DEVELOPMENT AND FUNCTIONAL ORGANIZATION The central nervous system (CNS) is unique in that its functions are carried out by, and thus depend on, the precise organization of its own intercellular communication systems or circuits. Study of the development and functional orga- nization of the neural circuits subserving cogni- tion, motivation, emotions, and other behavioral mechanisms is a major part of the research ef- fort toward understanding the etiology and treatment of mental disorders. Indeed, the ra- tional design and interpretation of behavioral and biochemical investigations of mental illness require a detailed knowledge of neural circuitry. Concepts about these systems, currently based on work done more than 20 years ago, can be expected to benefit from inclusion of the new insights into the normal control of behavior that will emerge from the application of modern pathway tracing techniques. The processes of neural development and functional organization are subject to the chem- ical and morphological influence of factors both intrinsic and extrinsic to the CNS. Understanding the influence of these factors is essential for complete elucidation of the neurobiological processes that underlie mental health and mental illness. The CNS plays a major role in regulating the chemical milieu and physiological conditions to which it in turn is exposed. This fact has two important implications for research in neuro- science: Not only must the significance of the effects of the chemical environment on brain function be determined experimentally, but the chemical products of the brain can often serve as convenient indices of brain function. The CNS is not fixed and immutable but capable of change and reordering throughout the lifespan of the organism. There is reason to believe that behav- ioral dysfunction may derive from altered de- velopmental patterning, synaptic reorganization, and age-induced changes in brain circuits. Neurogenesis and Development How cells decide which pathway of differ- entiation to follow, a central question in devel- opmental biology, is especially difficult to discover in the nervous system. Each neuronal phenotype is composed of a multitude of char- acteristics: the number, size, and shape of its axons and dendrites, the type and distribution of receptors, and so on. The question is often cast in terms of extrinsic influences vs. intrinsic predilections. That is, how much of a cell's fate is governed by its lineages, or family history, and how much is governed by its environment, ex- ternal influences, and cellular milieu? Knowing something about a neuron's capacity to adapt to changes in its environment could be useful in medical conditions where changing the balance of certain neurotransmitters is called for, or tissue grafts of a certain type are needed. Phenotype Control. Very little is known about the functional mechanisms that control neuronal characteristics, e.g., dendritic shape or the type of receptors present on the neuron's surface. Ina few cases, it has proved possible to control the pathway of differentiation taken by developing and even fully mature cells. Transplantation studies in embryos have shown that the devel-— opmental fate of neural crest cell populations is subject to local environmental signals; these cells can adopt any of several different pheno- types, depending on where the cells are placed. In the case of several neural crest derivatives, the nature of some of these environmental cues is becoming clear. For example, adrenergic sympathetic neurons can be converted into func- tional cholinergic neurons in culture by the ad— dition of a protein secreted by heart cells. Neuronal activity and certain hormones can also play a role in this transition. It is also possible to take undifferentiated cells and convert them into chromaffin cells or neurons, either adrenergic or cholinergic. Here again, the nature of the signals is becoming clearer. Recent work has also established that these potential transitions are, at least in one case, actually used in normal development. That is, neurons can and do assume different identi- ties during various stages of development. Advances in the area of phenotype de- termination must include studies to determine, in vivo, the role of the molecular signals that have been identified in vitro. This will require large-scale purifications and, in some cases, an- tibody production. Similar studies should be un- dertaken with CNS neurons. This will require work with cultures of dissociated cells from CNS areas of interest. Development of other pheno- ''typic characteristics such as receptors and sur- face antigens should be described, and attempts to manipulate their appearance should be undertaken. Cell—cell Recognition. Both membrane-bound and released macromolecules play major roles in cell-cell adhesive interactions that are critical to normal development of the nervous system. In addition to the intracellular molecules contained within the plasma membrane, there is a limited set of proteins and large carbohydrate-rich mol- ecules found outside of cells, including those associated with the external surface of the membrane (the cell coat) and those released by the cell as soluble molecules into the extra- cellular space. Aberrations in “neural-cell adhesion molecule" (N--CAM) have been associ- ated with one CNS lesion that is altered in a mouse mutant that has an abnormally con- structed cerebellum. Several proteins that appear to mediate some aspects of cellular recognition in neuronal tis-— sues have been isolated and well characterized. These include cognin and N-CAM, both of which have been isolated from the chick neural retina. Very large molecules that contain mostly car- bohydrate, called proteoglycans, are also inti- mately involved in the adhesive interactions of most cell types and are tightly associated with at least one of the adhesion-promoting molecules of the neural retina. Extracellular macromolecules also form more or less rigid matrices around cells and _ tissues. These matrices control many aspects of cell movement (e.g... nerve outgrowth) as well as_ the localization of neurotransmitter receptors to discrete areas of synaptic contact. A matrix protein has recently been isolated that apparently causes the clustering of acetylcholine receptors on muscle cells. The extracellular and membrane-bound mac- romolecules responsible for most of the cell—cell interactions within the nervous system may also modify the function of cells in their immediate environments. It will be profitable to continue efforts to isolate and characterize these molecules, using cultured cells as convenient assay systems, and to demonstrate their function in the intact animal. Trophic Factors. During early development, neuronal (and glial) cell behaviors are largely controlled by extrinsic microenvironmental agents. Similar controls are likely to apply to the functional maintenance of adult neurons, in- volutive processes in the aging nervous system, and development of at least some pathological situations. Extrinsic control may involve inhib-— itory influences, although only promoting ones have been identified thus far. Two of the several types of such factors that may be distinguished are surveyed here. Neuronotrophic agents regulate neuronal sur-— vival during development and neuronal mainte- nance or repair at more advanced life stages. Nerve growth factor (NGF), the first protein agent found to be trophic for certain sensory and sympathetic peripheral neurons, has prompted a search for other factors with demonstrable trophic activity on cultured neurons of the pe-— ripheral nervous system (PNS). Putative source materials were obtained largely from innervation territories or from glial cells. One such factor now purified involves cholinergic PNS neurons. More recent use of CNS neuronal cultures has revealed the occurrence of CNS-addressing protein factors as well as low molecular weight nonprotein agents. Neurite—promoting agents specifically direct the neuron to elaborate axons. Neurite—promot-— ing agents can act from the humoral environ-— ment of a neuritic growth cone (e.g., NGF itself), or they can occur in a bound form on cell surfaces and/or extracellular matrices. In the latter category, several new factors are now recognized and, in some _ cases, partially purified. These are: (i) very large, acidic molecules with both carbohydrate and protein moieties, (ii) released into culture media by several types of cells, including glia, (iii) unable to support neuronal survival by themselves but (iv) capable of vigorously promoting neuritic growth from both PNS and CNS _ neurons anchored to a suitable substratum. Neuronal survival in both PNS and CNS culture may require other trophic agents such as hor-— mones (insulin is a common requirement by all neurons examined in vitro) and also be regulated by certain neurotransmitters, ions, selected nu- trients, and other molecules related to neural activity. Survival requirements vary for neurons from different neural sources and/or of different developmental ages. Moreover, trophic sources are themselves likely to be under extrinsic regulation. Yet to be elucidated are the molecular mechanisms by which these various trophic agents operate, as well as the sequential or convergent ways by which they relate to one another. PNS and CNS in vivo models have been developed, which demonstrate that (i) trophic factors also occur in nerve and brain tissue of the adult animal, (ii) their levels are affected by neural lesions in both the damaged tissue and the fluids accumulating around it, and (iii) survival of neurons grafted into the lesion correlates well with the local accumulation of trophic agents. With respect to trophic factors, future re— search directions should encompass: e Purification and large-scale preparation of additional trophic and neurite—promoting agents, particularly those affecting CNS neurons, and generation of antibodies against them ''@ Elucidation of mechanisms of action for different agents on the same or on different responsive neurons e@ Development of cultures of adult neurons and definition of their changing require- ments for survival and neurite extension or maintenance e In vitro investigation of glial cells, their requirements, susceptibilities to extrinsic regulation, and abilities to provide (or mod- ify) their support to cultured neurons © Development of in vivo models to validate physiological roles for trophic and neurite— promoting agents identified in vitro and to probe their involvements in pathological states and/or repair processes Neuronal Death. This is a normal develop- mental phenomenon in which massive numbers of differentiating neurons degenerate. It occurs in all types of CNS and PNS neurons that have been examined, and it has been observed in all major vertebrate classes. Neurons that will ultimately die initially differentiate normally, including axonal growth and target innervation. Accentu— ated cell loss following target removal and re— duced cell loss following target enlargement provide compelling evidence for the important role ascribed to the target in regulating the survival of motor neurons. Afferents may also exert an influence on neuronal survival, since a substantial deafferentation increases the amount of neuronal loss. Neurotrophic factors supplied by either (or both) targets and afferents are probably involved in the regulation of neuronal survival. However, with the exception of NGF, no other neurotrophic factors have yet been isolated and shown to affect neuronal survival in vivo. Synaptic transmission is also involved in the survival of at least some types of neurons. Blockade of neuromuscular activity during the period of cell death prevents natural motor neuron death. Synaptic activity may regulate the availability of a target-derived trophic factor (i.e., inactivity could increase the availability of such a factor). The prevention of cell death by neuromuscular blockade is associated with an increase in acetylcholine receptors in the mus-— cles of these preparations. Blockade also induces an increased arborization of motor neuron axons and terminals, resulting in a hyperinnervation of these muscles. The biological value of cell death remains to be understood. It appears that cell death is not an error mechanism for removing aberrant ax- onal connections. The overproduction and sub- sequent loss of neurons may be a mechanism for quantitatively matching neurons with synaptic targets. Embryonic neurons may compete for a target- or afferent-derived trophic factor that is in limited supply; the losers in this competi- tion would be the cells that undergo natural cell death. Hormonal Requirements. The requirement for thyroxin in normal neuronal development is a clear example of the powerful effect hormones have on neuronal development and function. The list of hormonal and chemical factors that affect CNS structure and function, transiently and permanently, is already lengthy and most likely incomplete. Such factors include hormones with generalized metabolic actions (e.g., thyroxin, insulin, growth hormone, adrenal steroids), fac- tors appearing to have more specific actions (e.g., gonadal steroids, adrenocorticotropic hor- mone (ACTH), the multiple gastrointestinal hormones), and factors of neural origin (e.g., neuropeptides). Certain generalizations can be made: (i) Brief exposure of the nervous system to hormonal factors during critical periods of de- velopment can lead to permanent changes in brain structure and function. (ii) Physiological fluctuations in hormonal secretion during adult life modify the function and probably the structure of the nervous system. (iii) Hormonal factors (adrenal and gonadal steroids) seem to play a major role in various dysfunctions of the aging brain. As an example of these phenoména, gonadal hormones will be discussed here, but it must be emphasized that the modification of neural development by hormonal factors other than these may underlie the development of mental illnesses (see chapter III). The gonadal steroid environment permanently alters the functional capacity of the mammalian CNS: Functional sex differences are imposed on what appears to be an inherently female, or possibly neuter, brain by the action of testicular hormones in the male. Recently, structural sex differences in the mammalian CNS have been identified and shown to be controlled, perhaps completely, by the hormone environment. They range from subtle differences in synaptic ter— minations, dendritic arborization, and neuro- transmitter input, to relatively gross differences in regional nuclear size in the hypothalamus and surrounding regions of the brain and in the spinal cord. It now appears that steroid hormones may be neuronotrophic substances, either by directly promoting the survival of specific hormone-sen- sitive neurons, or doing so indirectly by the stimulation of the growth of neuronal connec— tions. Moreover, it still has not been ruled out that the gonadal steroid environment can modu- late such fundamental processes as neurogenesis, neural migration, and perhaps cell-cell recog- nition during the organization of the CNS. Neural Transplants and Plasticity. The role of neuronal activity in cellular interactions during development has been neglected in the study of ''neurogenesis. A plausible hypothesis is that the same electrical and chemical signals that the brain uses in its mature functioning may also help guide its development. Early appearing action potentials, slow potential changes, transmembrane ionic currents, neurotransmit- ters, synaptic transmission, and sensory input are all candidates for developmentally relevant signals. The advent of new technology in neuro- science makes this a propitious time to focus research on questions of neural plasticity. Since many aspects of neurobehavioral plasticity probably involve changes in the physiology of specific neuronal networks, an understanding of how neuronal activity regulates normal devel-— opmental events should also aid our under- standing of plastic mechanisms. What are the ways in which the nervous sys— tem can repair itself or be influenced to reorganize in a manner that leads to restoration of function? New knowledge may yield insights into a number of basic processes, notably in the areas of cell interactions and trophic influences, and may bear implications for psychiatric and neurologic diseases. In particular, strategies involving the transplantation and reconnection of specific groups of nerve cells may be used to investigate certain disorders in animal models of disease (e.g., septohippocampal transplants and memory). Symptoms displayed by models for several disease states that result from neuronal loss have been ameliorated by intracranial brain transplants. For example, diabetes insipidus in the vasopressin-deficient Brattleboro rat has been improved by hypothalamic implants, as has hypogonadism in mice deficient in gonadotropin releasing hormone. In both human and animal models, Parkinson's disease is being treated with adrenal implants into the CNS. The current state of transplant art is based on data from a wide variety of studies. A new ap- proach used in mammals is to employ transplants to (i) replace neuronal populations and (ii) in- vestigate the potential for axonal regrowth from indigenous cells in the adult brain and spinal cord. With respect to the first objective, neu-— ronal transplants usually consist of fetal or new- born nerve cells grafted into the adult host neuroaxis, where they survive, differentiate, and grow processes. In some cases, the implanted cells have been shown to improve or correct functional deficits in experimentally injured or mutant animals. With respect to the second ob- jective, although it was once held that fibers fail to regrow after they are severed in the CNS of adult mammals, it now appears that nerve cells in many regions of the brain and spinal cord re— tain their capacity for renewed axon growth. When nonneuronal components of the CNS are substituted by those from the PNS, axons from central neurons can successfully elongate again for distances equivalent to those of the long projection and association tracts. These regen- erated CNS neurons retain many of their natural functional attributes. It is not yet known whether the neurons iden— tified in these experiments represent a special group or are examples of a more general regen— erative capacity. It also remains to be estab-— lished whether these cells make functional synaptic connections with the cells they reach. The nature of the influences that limit axonal growth in the CNS and permit or promote elon— gation along these transplanted PNS conduits is also unknown. Inhibitory and stimulatory influ ences arising from these nonneuronal compo- nents may depend on surface properties of cells, their spatial arrangement, the release of specific molecules, or the presence of certain critical components of the extracellular matrix. In sum, however, these studies now permit exploration of the elements forming a neuronal circuit. Future directions for transplantation strat- egies include investigation of intrinsic neuronal mechanisms influencing cell survival and growth and of anatomical and functional connectivity between regenerated nerve fibers and their targets. In particular, there is a need to know if functional effects are determined transsynap- tically or as a result of the release of trans-— mitters into the surroundings of denervated cells. It will also be important to determine the nature of the interactions between nerve cells and their immediate environment that influence neurite outgrowth and connectivity. The success of attempts aimed at neuronal circuitry recon— struction will depend on understanding the cues that guide axons over short and long distances to their targets and further characterization of the functional properties of regenerated and rein- nervated neurons. Reported sex differences in the plastic re— sponse of the CNS to lesions in the adult may be yet another manifestation of the perinatal de- velopmental effects of steroids. However, go- nadal hormones probably modify the plasticity of the CNS throughout life. In the songbird, for example, gonadal hormones appear to exert seasonal influence on regional nuclear volume, dendritic growth, and synaptogenesis. An ap-— preciation of the dynamic structural plasticity of the adult brain, and its probable control by go- nadal hormones, has important implications for an understanding of mental illness. An important direction for the future is to illuminate the mechanisms by which hormones determine the structure of the CNS. Such studies should include those aimed at identification of other structural sex differences. A specific morphological signature of hormone action may provide the most productive model for future studies at the molecular and genetic level. The hypothalamic sex difference in the rat, i.e., the sexually dimorphic nucleus of the preoptic area, serves as a model for the study of the influence of gonadal hormones in turning neuronal genes on ''and off during development. This model system may be amenable to monoclonal antibody and recombinant deoxyribonucleic acid (DNA) tech- nology. Also needed are studies aimed at descriptions of the anatomical and chemical connectivity of sexually dimorphic CNS regions, their development, and the influences of experimental manipulation of the hormone environment. Aging. The aging brain represents a condition in which long-term, gradual changes in neural structure, physiology, and chemistry lead to de- clines in learning, memory, and other aspects of adaptive behavior. The condition of senile de- mentia of the Alzheimer's type appears to in- volve a dramatic acceleration of such changes. Elucidation of the mechanisms that underlie this gradual reorganization should yield an increased understanding of the processes governing neural maintenance, reorganization, and plasticity. Thus, the aging brain is an important experiment of nature that offers a powerful analytical tool. A common corollary of aging is reproductive senescence, particularly in the female. Consid- erable effort has gone into the study of repro- ductive senescence, especially ovulatory failure, as a model of hypothalamic aging. One com- monly proposed theory is that ovarian hormones exert deleterious effects on the adult hypo- thalamus; these effects could be neurochemical and/or morphological. In any case, it is clear that ovariectomy can delay or reverse the symptoms accompanying reproductive senes- cence, whereas treatment with exogenous steroids, particularly when administered peri- natally, appears to lead to precocious repro- ductive senescence. Adrenal steroids have also been implicated as a potential factor in neural aging, especially in the hippocampus. Studies of the role of hormones in the neuronal aging process should address the following Key problems. What is the nature of the mechanisms that chronically modulate brain organization during aging? What are the synaptic and other neural mechanisms that specifically underlie behavioral—cognitive changes? Which changes are reversible and which irreversible? Neural Circuitry Advances in methodology now permit the analysis of functionally relevant neural systems, with unprecedented specificity and sensitivity, at a number of levels of organization. These ad- vances include: e Pathway tracing methods--autoradiography of radioactive amino acids, horseradish peroxidase, and other protein markers ® Immunohistochemistry and in situ hybrid- ization— identification and localization of transmitters/modulators, proteins, glyco- lipids, receptors, and preparation of mono- clonal antibodies for identification of unknown molecules and for the study of transmitter specific pathways, applications to transmission and scanning electron microscopy @ Receptor binding autoradiography- -quali- tative and quantitative studies of specific localization ® Computer applications to quantitative anatomy--e.g., autoradiography, immuno- histochemistry, and serial reconstruction of electron micrographs—and to three-dimen-— sional graphics display of brain connections, typology, transmitters, etc. e Microscopy- transmission, scanning, scan-— ning-transmission, and X-ray electron mi- croscopy, and proton—beam microscopy e Freeze fracture studies of membrane par- ticles (receptors, synapse specific proteins, etc.) and determinants of their insertion, aggregation, degradation, and responses to drugs e In vivo imaging - positron emission trans- axial tomography, single photon emission tomography, and nuclear magnetic resonance These new methods are providing highly re- fined tools for the study of the nervous system. Such research, discussed later, requires the use of neuroanatomical and neurophysiological techniques applied at the cellular, multicellular, and behavioral levels. Each approach has con- tributed to knowledge of normal and abnormal neural organization. An important approach of interest in anatom- ical-behavioral analysis is neuroethology. Study of the circuits involved in naturalistic “units of behavior" has increased with the advent of so- phisticated technology permitting refined ex- perimentation. As noted, study of vocalization in songbirds has revealed endocrine regulation of both cell number and dendritic outgrowth in specific brain regions. This work has also yielded a clear-cut demonstration of lateralization of vocal functions. This may be analogous to the mechanisms mediating mammalian vocalization, although a suitable mammalian model is needed to investigate the possibility. A fundamental concept underlying the neuroethological ap- proach is the evolutionary conservation of neural organization across species. Fundamental units of behavior are generally found to be mediated by homologous neurons. Insight into the cellular and molecular mechanisms underlying evolu- ''tionary change should provide fundamental un- derstanding of the human nervous system. Other areas ripe for anatomical—behavioral analysis include: the maturation of the nervous system in relation to critical periods, in behavior development, and pharmacological manipulation. Studies of the cellular and molecular basis of drug action remain the areas of greatest activity and are discussed at length elsewhere (chapter IV). Neuropharmacology of the developing and mature brain is also of interest. Recent advances in electrophysiological in- vestigation of specific neurotransmitter and neuromodulatory actions have provided the means to examine the influence of these agents in well-defined preparations with great preci- sion. However, we are only beginning to under-— stand the physiological interactions among the various identified neurochemical circuits in the brain. A key objective in future investigations of neural organization is to establish not only how the various neurochemically defined anatomical systems operate, but also when they operate in the normal moment-to—moment and day-to-day activities of organisms that possess them. It is only after these qualifications have been met that the neural circuitry responsible for proc- esses such as complex cognitive functions will be understood at the cellular and synaptic level. All of these measurements need to be closely correlated with microvascular changes, blood flow, and changes in brain metabolic activity. Merging of information from extracellular fluid (ECF) dynamics with 2-deoxyglucose (2-DG) positron emission tomography (PET) scanning and similar approaches, discussed later, offers a possible inroad to better understanding of neural organization. Among other needs, it will be important to consider the influences of ECF and surrounding tissue elements on the properties of identified neuronal systems. Probes are now available that monitor chemical fluxes in the ECF and operate on a time scale consistent with ongoing activity both in isolated neural preparations and in the intact mammalian brain. Ion-selective micro- electrodes (ISMs) allow continuous monitoring of ionic fluxes, which are critically associated with important neural responses. Other ECF moni- toring probes have been developed that follow concentration changes of the catecholamine neurotransmitters, their metabolites, and related species. These voltametric electrodes, together with ISMs, complement and extend the range of electrophysiological data that can be obtained from functional analysis of specific neural sys- tems. Electrophysiologists with expertise in multibarrel pipette fabrication can build an array containing multiple ISMs, recording elec-— trodes, and electrochemical detectors to monitor the functional aspects of ionic and neurotrans— mitter fluxes. With these techniques, one can talk knowledgeably about the spatiotemporal distribution of chemical species in the ECF that may modulate neuronal activity. Important new directions in this area include understanding the modulatory characteristics of “slow acting" neuropeptides. Functional Mapping. In vivo autoradiographic imaging techniques, originally developed for the study of cerebral blood flow and metabolic reg- ulation, have been used to map the functional organization of specific CNS pathways. Recent- ly, the autoradiographic 2-DG method has been modified for functional mapping of the human brain. Yet another application of autoradi- ography holds promise for assessing local changes in brain protein synthesis; this could greatly enhance studies of brain development, degeneration, regeneration, and neural plas— ticity. In other work, the development of a com- puterized image-processing system has largely overcome the problems associated with the de- tection of differences in functional labeling in black-and-white autoradiograms. This imaging technique generates pseudocolor-coded maps depicting the distribution rates of local cerebral glucose utilization. These various global mapping techniques, coupled with image—processing techniques, have been invaluable in bridging the gaps between detailed anatomical and electro- physiological studies and behavioral studies as well as between the results obtained in animal and human studies. Nonautoradiographic imaging techniques such as the flavoprotein fluorescence, voltage sensitive dyes, and nuclear magnetic resonance techniques have yet to be validated for global, behavior—related studies. The following basic types of 2-DG studies have been reported: (i) methodological studies to determine the resolution-limiting factors of the technique and to modify it for use at the cellular level; (ii) mapping studies of established path- ways to answer questions about specific types of functional activation; (iii) functional plasticity studies related both to normal development and to effects of sensory enrichment; (iv) pharmacological studies on the effects of various anesthetics, psychotomimetics, and transmitter agonists and antagonists; (v) pathophysiological studies of metabolic changes associated with experimentally produced seizures and hypox-— emia; and (vi) behavioral studies, which include examination of the effects of self-stimulation, hibernation, circadian rhythms, and movement. The paucity of behavioral studies to date is un— doubtedly the result of technical limitations imposed by the quantitative 2-DG method. This limitation is being overcome by employing the method semiquantitatively. In this case, inves-— tigators dispense with the arterial catheters, inject the 2-DG via either a chronically im- planted venous catheter or intraperitoneally, and compare changes in radioisotope concentrations or relative optical densities on the autoradio- ''grams. The 2-DG method will continue to be applied in its present form to the functional mapping of neural circuits, both classical and more complex, and/or circuits related to specific complex be- haviors, e.g., motivation and affect, learning and other cognitive behaviors, and complex motor acts. With improved resolution and the refine- ment of double-label methodology, these studies will now involve a more detailed examination of changes in functional labeling produced by subtle changes in such stimulus parameters as fre— quency, amplitude, and orientation. The exact localization of 2-DG labeling within the neuropil, presently unknown, will require continuing development of the tritium-2—DG method. This will also permit a more detailed analysis of functional metabolic changes at the cellular level. In addition to the special histo— logical techniques (freeze-—substitution, etc.) usually required of water-soluble labels, new approaches such as the use of a specific deoxy- glucose-binding lectin could eliminate the need for special tissue preparation techniques. Con- tinued development of the semiquantitative 2-DG method is of particular significance for the study of behavior in freely moving animals, as is the development of 2-DG PET technology for mapping function in the brains of normal and mentally ill subjects. Biochemical Analysis. A thorough knowledge of functional neuroanatomy is prerequisite to understanding neural functions in both health and disease. In some ways, it might be argued that the entire existence and structure of the neuron are focused on delivering its neurotransmitter to the right place at the right time. Beyond the question of what neurotransmitter is used by which cell is the issue of how to apply that knowledge to understanding brain function and treating CNS illnesses. Contemporary psychiatry obtains its major therapeutic tools from agents that alter neuro— transmission. It is likely that our ability to un- derstand and treat mental illness will grow in proportion to our knowledge of the biochemical subtleties of neurotransmission. The question is—Why perform these studies with anatomical tools, when the biochemistry suffers because of tissue preparation? Why not simply study the biochemistry in fresh, unaltered blocks of brain? The answer is twofold. First, the biochemistry must. indeed be done with as much precision and depth as possible. The second answer is equally clear: The gross tissue type of biochemistry (e.g., a 20-mg block of hypothalamus) clearly distorts and loses the very information that the brain spent so much energy transmitting. Ana- tomical tools impart the ability to discover both key organizational principles and levels at which we might manipulate disease states. Immunocytochemistry has been a rich source 10 of information on neurotransmitter substances and their biosynthetic enzymes. These studies have not only elucidated transmitter biosyn-— thesis, they have identified a new group of sub- stances to be localized, mapped, and understood. There are now more than SO peptide neuro— transmitter/neuromodulator candidates, which should generate much research in primary map— ping, not to mention specific connectivity studies. Yet the problem is even more complex than the candidacy of 50 peptides would imply. In the course of anatomical and biosynthetic studies on the same substances, it was recognized that many cells used combinations of biologically active substances for neurotransmission. Demon- stration of cotransmission has its own technical problems, but these were substantially com-— pounded when families of peptides with different precursors but similar biologically active cores (e.g., the opioids) were discovered. Hence, it has become clear that several types of technical controls are essential before meaningful, ana- tomically and biochemically accurate informa-— tion can be obtained from immunocytochemistry studies. Further, immunocytochemistry must be coupled with protein biochemistry and molecular biology in order to resolve the serious specificity and cross-reactivity problems. In conjunction with anatomical studies, mi- crochemical assay methods are particularly suited to quantitative studies of the concentra— tion gradients of neuromodulators. This approach to neurotransmitter distribution patterns can be applied to the study of human postmortem ma-— terial. In so doing, pathophysiological correlates of certain psychiatric diseases may be discovered. The most recently acquired tool of biochem- ical anatomy, in situ hybridization, uses com- plementary DNA (cDNA) probes to visualize specific messenger ribonucleic acid (mRNA) species in neuronal and endocrine tissue. The method integrates tools of the molecular biol— ogist with those of the histochemist. The goal is the visualization of a specific mRNA in a par-— ticular cell type in the CNS. To date, the main focus has been on neuropeptide precursor mRNAs, but there are growing efforts at lo- calizing the mRNAs for synthetic/degradative enzymes, receptors, and general CNS proteins. The method has several strengths. For example, one might ask if there are changes in beta-— endorphin/ACTH mRNA under conditions of stress, in which a substantial peptide release is seen. In other conditions, where peptide con- centrations are stable but turnover is increased, the mRNA may well be a good window on bio-— synthetic rate and cellular activity. A second advantage includes single-cell resolution, es— pecially useful when several cell groups in the same area use the same transmitter but for different purposes. A third advantage stems ''from the nature of a DNA--RNA hybrid. The de- gree of stringency of the match can be con- trolled, thereby allowing estimation of degree of sequence homology or available length of mRNA. In immunocytochemistry, future directions point to the mapping for circuits of specific substances. Relevant questions include the fol- lowing: (i) Where is it? (ii) Which cells and fibers are connected? (iii) Which neurotransmitter systems "talk" to each other? (iv) Which sub- stances does the neuron package as neuro- transmitters? Immunocytochemical studies should also explore the biochemical and techni- cal aspects of neurotransmitter systems; for example, what are the specific forms of peptides in particular cell groups? While a particular ge- nomic precursor may be known, the cleavage product and posttranslational forms frequently are not, especially in different cell groups. The question here is: What is the cell actually using for transmission? Finally, development of crit- ical immunocytochemical components will be essential, i.e., synthesis of key peptides, puri- fication of key enzymes, production of specific antisera, and use of monoclonal antibodies. With respect to in situ hybridization of spe- cific mRNAs by cDNA probes, certain technical issues are of paramount interest: (i) covisual— ization of protein and mRNA in the same cell, an essential validation step, especially early in the development of a technique, and (ii) estimation of the change in mRNA levels using in situ methods. Studies of the effects of pharmaco- logical and behavioral manipulations on mRNA levels via densitometry or grain counting should be coupled to mRNA quantitation. Both of the preceding anatomical methods, immunocytochemistry and in situ hybridization, offer the neuroscientist a unique window into neurotransmitter biology. But the strengths of these methods do not stop there. They offer the potential for studying cellular neurotransmitter biochemistry. One of the most difficult areas of CNS study is that associated with the dynamics and regulation of specific neurons. What causes a particular cell to fire? What do antagonists do to its firing and release rate? Is it increasing its synthesis and release of neurotransmitter in a particular illness? These kinetic issues can be studied only poorly in tissue blocks, which mix the biology of large numbers of cell types. In contrast, the new biochemical and anatomical tools already described provide the methods and resolution to begin to study this regulation problem in vivo. For example, one can ask if a particular mRNA ACTH species is increased when the organism is under stress. If so, it strongly implies increased synthesis and release of the ACTH peptide. Or one might ask if tyro- sine hydroxylase is found in greater abundance in a particular subset of dopamine cells in schizo-— phrenic brains. The number of uses of the higher resolution biochemical—anatomical tools is large; 11 although there are substantial problems, they offer real hope for improving our knowledge of the CNS and its diseases. Receptor Localization. Different classes of receptors determine how chemical mediators such as neurotransmitters and hormones influ- ence the excitability of cells. Neurotransmitter receptors are defined by their pharmacology as demonstrated in receptor binding studies. Radi- oactively labeled ligands can be used to tag these receptors in vitro, and the kinetics and pharmacology of this ligandreceptor interaction used to define the receptors. Such approaches have been used to label CNS receptors for a variety of neurotransmitters, including norepi- nephrine, acetylcholine, dopamine, serotonin, gamma-aminobutyric acid (GABA), histamine, and several opioid and non- opioid peptides (see chapter II). One emerging principle is that many transmitters interact with more than a single class of receptor, and that these receptor sub- types may be differentially distributed in the brain. Although receptor location may be studied at the gross level by homogenizing brain regions and determining ligand-receptor interactions in the resultant fractions, more precise localization requires their identification in slices of brain tissue. The most satisfactory current approach involves sectioning frozen brain tissue, incu-— bating the slide-mounted sections in low con- centrations of a radio—labeled compound known to bind in a saturable manner to the site of in- terest, washing the sections, and then exposing them to a radioactive-sensitive emulsion and developing that emulsion. In these experiments, the kinetics of the association and dissociation of the labeled ligand can be determined, as can the optimal incubation conditions and the phar- macological definition of the site to which the labeled ligand binds. None of these ligands bind exclusively to the site of interest in brain mem— branes, but those that are useful in autoradi- ography are at least 60—percent specific. In many experiments, it is important to achieve not just a qualitative picture of the CNS structures to which the ligand binds, but also the quantity of binding in the presence of saturating concentrations of ligand. This affords an indi- cation of the number of receptors present in a given brain region. Quantification can be achieved using tritium-sensitive film. After de- velopment, the films can be read for optical density using a densitometer or video camera. When radioactive standards are included with each piece of film, conversion of optical density values to amount of ligand bound is feasible. More recent advances in the quantification process involve using a video camera to collect the autoradiographic image and then, using a calibration curve based upon the standards, ex-— pressing the grey value of each pixel in that ''image as a linear function of the concentration of ligand bound in that area of tissue. Receptor gene cloning is another approach to study the function of receptors. Final proof that a ligand binding to a putative receptor protein is of functional significance may only emerge in studies that clone the cDNAs or genes coding for the putative receptive proteins. This was re- cently accomplished for the nicotinic acetyl- choline receptor, and the amino acid sequence for this receptor is also now known. In addition, DNA probes are now available to study the reg- ulation of this receptor. Cloning the genes for other important receptors is now underway in a number of laboratories. Gene cloning provides a powerful approach to the sequencing of neuro-— transmitter receptors. Knowing the sequence for a receptor makes it possible, in turn, to raise specific antibodies for immunohistochemistry and purification and as probes to study receptors in human brains. DNA probes also can be used to measure mRNA levels and to study the regula- tion of expression of receptor genes during nor- mal development, after experimental manipula- tion, and in various disease states. Future directions for approaches to receptor localization should address (i) the relationship between the postsynaptic receptor and the pre- synaptic afferent innervation, (ii) up— and down- regulation of receptor number as determined by neuronal activity, and (iii) drug-—or environ- mentally induced changes in the number and lo- calization of receptors that may contribute to the overall functioning of neural circuits in- volved in behavioral processes. These problems may be approached through either ligand binding or, when available, gene cloning of receptors. Finally, the area is poised to begin serious quantitative examination of the regional dis- tribution of various receptor subtypes (cholin— ergic, dopaminergic, noradrenergic, GABAergic, peptidergic) in the postmortem brains of patients with psychiatric disorders. Needed are large blocks of morphologically intact frozen tissue made available to investigators with expertise in particular receptors. Modification of Brain Development and Function Receptive Mechanisms. The nervous system is continously affected by a variety of such agents as neurotransmitters, trophic factors, toxins, proteins, peptide and steroid hormones, metabo-— lites, and viruses. Their influences are respon— sible for short- and long-term alterations in the structure and function of the nervous system. Influences are perceived by neurons via recep— tive mechanisms located on or within a variety of cellular organelles. Three mechanisms for delivering influences to target cells are recog- 12 nized: (i) strictly private one-to-one communi-— cation of cells through synaptic contacts and/or intercellular pores, (ii) local systems serving multiple adjacent target cells, that is, paracrine systems, and (iii) general delivery systems, e.g., the bloodstream, which transmit signals to all parts of the body. Influences are perceived by the appropriate target and ignored by inappro- priate ones by means of receptors. Currently, a number of receptive mechanisms are recognized and have proved useful in explaining normal and pathological states as— sociated with environmental and hormonal in- fluences. More scope and flexibility is needed in studies in this area. For example, it is dogma that steroid hormones act via cytoplasmic re- ceptors (or receptors loosely associated with the nucleus), which translocate to or become tightly associated with chromatin in the nuclear com- partment, and that, within the nucleus, recep-— tor-steroid complexes mediate the alteration of genomic expression in target cells via activation of DNA-dependent RNA polymerases and ex-— posure of new and/or additional DNA sequences for transcription. This view of steroid action cannot account for the variety of known steroid effects and precludes alternate receptive mech-— anisms at the cytoplasmic, membrane, or ex-— tracellular level. Important questions to be addressed in the near future include the following. How is the delivery of environmental and humoral in- fluences regulated? Important regulatory mech- anisms that should be studied include the blood-brain barrier, specific neuronal and glial transport mechanisms, the role of serum-binding proteins in humoral delivery, and hepatic and renal metabolism and excretion. How are these external influences perceived by the appropriate target and ignored by inappropriate ones? Re- ceptive mechanisms other than those currently recognized should be investigated. Active searches should be initiated, for instance, for membrane-bound steroid receptors, extracellular transport and receptive systems analogous to low-density lipoprotein for cholesterol, and nuclear receptor-acceptor mechanisms for neurotransmitters. Hormonal Influences. The hypothalamic reg- ulation of the autonomic nervous system directly links the CNS to the peripheral chemical envi- ronment, which includes circulating levels of epinephrine and the various gastrointestinal hormones. Moreover, the direct innervation of peripheral endocrine glands such as the adrenal cortex, pancreatic islets, and gonads may reg- ulate the hormone environment, if not by direct secretomotor function, then by a possible al-— teration of blood flow to these organs. Certain neurochemical factors may be of physiological significance entirely within the CNS, since the various circumventricular organs may represent ''sites of chemical access to the cerebrospinal fluid. There is a clear need, therefore, for studies of the neural control of hormonal activity, in- cluding the biosynthesis, processing, storage, and release of these chemical factors, and of the impact of these substances on the control of brain function, including the identification and localization of hormone-producing and hormone- responsive neurons and the mechanisms of hor- mone recognition, processing, and action. Of particular interest are the neural mechanisms underlying the physiological responses to stress. These responses produce dramatic changes in the immune and endocrine systems, which in turn influence brain function and behavior (see chap-— ter III). In addition, neurotoxins and infectious agents, for example, should be studied in the context of their impact on brain structure and function (see chapter IV). Several factors completely extrinsic to the organism can also influence brain function either directly or via modification of the neuroendo- crine axis. Circadian Circuitry. Photoperiod is a factor of great importance. Studies of the chronobiology of all aspects of brain function, including the response to external rhythms and the generation and chemical mediation of internal rhythms, are necessary to understand brain function in the temporal domain (see chapter III). The outline of the vertebrate “circadian axis" has become clearer in recent years. This axis includes the retina, pineal gland, and suprachiasmatic nucleus (SCN) as well as the links among these struc— tures. Some of these links are well known (e.g., the retinohypothalamic tract and sympathetic innervation of the pineal), while others remain unexplored (the route by which pineal hormones affect the SCN). It is now important to study the axis dynamically, asking questions about its or- ganizational structure, the flow of information among its components, and the effects of en- dogenous and exogenous perturbations. Answers to these basic questions will certainly be rele- vant to the treatment of pathological states that are either caused by disruption of the circadian system or influenced by that system (see chapter III). Nutrition and Diet. Nutrition, environment, and other experiential factors that ultimately affect the brain are important to understand in terms of their influence on developmental plasticity. In the past, work on nutrition and diet focused on the impact on brain function of such nutritional deficiencies as chronic protein and/or calorie malnutrition. Effects on brain were re-— ported, but they were inconsistent and undra-— matic. Recently, the normal vagaries of nutri- tion have come under study, leading to a focus on the effects of variations in normal diet (e.g., 13 how much protein is consumed during each meal) on brain chemistry and particular brain functions. Of necessity, the connection of diet to brain function must include such intervening variables as (i) blood levels of particular nutrients and how diet-related variations in them affect the nu- trition of the brain, (ii) blood levels of particular hormones and how or whether diet-related changes in their levels modify brain functions, and (iii) blood-brain barrier transport systems for many nutrients and how their functioning is influenced by diet and other variables. A clear corollary line of investigation concerns the ef- fects of specific metabolic diseases—-e.g., diabetes—on the blood level of these same nu- trients and hormones, on their brain transport carriers, and ultimately on brain chemistry and function. If the nutrition—brain function area is broadly defined in these terms, then one may note many nascent lines of investigation suggesting interesting connections. Two examples follow. Ingestion of single meals by rats has been found to modify blood levels of tryptophan, ty— rosine, and several other large neutral amino acids. These changes lead to predictable changes in the uptake of tryptophan and tyrosine by brain neurons and, as a consequence, in the rates of their conversion to particular neurotransmitters. Some data now suggest that changes in trans-— mitter formation induced by diet infuence specific brain functions and behaviors—e.g., the regulation of appetite, blood pressure, the se- cretion of growth hormone and prolactin, and the quality of sleep. In humans, the importance of dietary intake is exemplified by the metabolic disease phenylke— tonuria. In this condition the body cannot me- tabolize the amino acid phenylalanine at the normal rate or to normal breakdown products, causing profound mental retardation. The de- structive effects on brain chemistry and function can be largely prevented—-if the condition is de- tected early after birth—-by rigidly limiting di- etary phenylalanine intake. While many chemical hypotheses on the effect of high blood phenyl- alanine on brain chemistry and function have been developed and tested in animals, none has withstood the test of time. Since dietary control of this disease is not always possible or optimal, the discovery of how high blood levels of phenylalanine cause mental retardation is an important research goal that could lead to the development of better treatment modalities. Several general questions can be identified for future investigation in the area of nutrition— metabolism—brain function. These include the following. e Which specific nutrients influence brain chemistry and function, how, and under what conditions? More broadly, in response ''to changes in nutrient intake, and/or the secretion of metabolic hormones in health or disease, what specific changes occur in brain function and behavior? For example, does the brain specifically alter (i) meta— bolic and/or hormonal parameters in the periphery and/or (ii) the appetite for and/or the intake of specific nutrients in a manner consistent with an attempt to control nu- trient fluxes in the body? Are there other behavioral effects? ® Do the hormones that are released in as— sociation with nutrient ingestion influence brain chemistry and/or brain function, and how? This issue involves determining if hormones released on food ingestion, or during changes in prandial state (e.g., during fasting), influence the brain chemically, either directly or indirectly. e If nutrients and hormones normally influ- ence brain chemistry and function, whether directly or indirectly, are brain chemistry and function adversely affected in a pre- dictable way when specific metabolic and/or genetic diseases render the disposition of nutrients and the release of hormones ab- normal in response to food ingestion? This question includes the issue of whether the metabolic and hormonal abnormalities as- sociated with diabetes affect brain chem-— istry and function via altering nutrient supply to the brain (e.g., amino acids, glu- cose, ketone bodies). Experience. It is now known that experience is a major determinant of the organization and function of the nervous system in higher mam- mals. Recently, characteristics of the rearing environment have been shown to have profound effects on brain and behavioral phenotype, and the brain continues to be molded by experience as adult behavior is changed. This view has de- veloped gradually from a behavioral perspective; the knowledge that these effects are presumably underlain by changes in the number and structure of the synaptic connections between nerve cells has emerged in the past 3 to S years. Most of the evidence leading to these conclusions has come from three basic paradigms—-sensory/social deprivation and enrichment, formal training in adult animals, and electrical stimulation. The changes affect not just the organism's behavioral patterns and capacity but also appear to be tied to its ability to resist later environmental prov— ocations leading to stress, disease, seizures, and other detrimental events. Competition between afferents exists during development and probably throughout life. Ini- tially, in development, connections are over- produced. As a result of experience, it appears that a subset of these is stabilized, and the ex- 14 tras degenerate. The losses are massive and rep- resent a considerable commitment of metabolic resources, which suggests that the mechanism provides advantages over other means of achiev— ing neural organization. Synapse formation is not restricted to devel- opmental periods as traditionally conceived. ° Synapses form in adults as a result of damage to competitive systems, electrically stimulated activity, and behavioral experience, including formal training in the adult. This may be one mechanism whereby brain organization is con- tinuously tuned to experience; the mechanism could underlie aspects of memory. Experiential modulation of synapse formation is not restricted to developing sensory systems. Plasticity occurs not only in sensory-motor re- gions (cerebellum, motor cortex) but also in less peripherally tied regions (e.g., hippocampus) and may, at least in development, be the rule rather than the exception for neural tissue. Under at least some conditions (e.g., electrical stimulation—produced organization changes, as in long-term potentiation), structurally recogniz— able synapses appear to form within minutes in the adult brain. The generality of this finding outside of the hippocampus has not been demon- strated, although physiological findings in other systems are interpretable in these terms. A primary area in which great strides have been made is the study of neural plasticity at the anatomical level. While we know that more ex— perienced animals have more synapses in some areas and perhaps different patterns as well, we do not know how experience causes this nor how new or selectively stabilized connections func— tion in behavior. Moreover, we do not know how or whether the overproduction—loss mechanism is general or restricted to the few, primarily sen— sory systems studied. Finally, in the same ex— perience manipulation paradigms, synapses typ- ically have altered structural characteristics. We do not know whether these structural charac- teristics merely reflect the stabilization process or whether they represent parallel changes in the strength or other operating characteristics of these neuronal connections. Turning to future research on the effects of experience, combined electrophysiological, molecular, and morphological studies of the mechanisms whereby experience and activity stabilize CNS connections are necessary. We need to know whether experience/activity can generate connections and, if so, how connections are generated in patterns appropriate to brain organization. We need to know whether changes in the structure of synapses brought about by experience/activity are related to their stabi- lization or whether these changes are associated with a separate type of functional change. To pursue these issues, we must improve ways of data collection in parallel from whole—neuron microscopic methods, such as Golgi procedures, ''and methods allowing focus upon individual syn— apses, such as conventional and novel electron microscope methods. Use of stereological elec— tron microscopic analyses, now developing, of— fers real promise in this regard. In addition, "dynamic" neuroanatomical procedures, in which the action of a peripheral synapse can be viewed in the midst of milli- to microsecond steps of the transmission process, must be extended to the CNS. Nonsynaptic forms of interneuronal communication should also be examined, as should reported “synapse splitting" and other mechanisms of synapse formation that fall outside traditional concepts. Broader aspects of experiential effects on the nervous system should also be vigorously in-— vestigated. Most of our knowledge of sensory system function and effects of experience thereon comes from studies of the visual system because of its accessibility and ease of manipu- lation. Study of other sensory systems is neces— sary to determine whether the principles developed in vision research are general, or whether other rules apply that may modify our clinical approach to these systems and our range of neural experience—encoding mechanisms. A particularly exciting phenomenon to be un- derstood is the appearance of large numbers of mRNAs (gene copies expressed in cells), which are unique to the brain and appear gradually from some time before birth to adulthood. Many of these genes are probably involved in the unique features of the many different types of neurons in the brain, but some may be involved in the processes of brain organization, and mod- ification by experience. Simple tests of the ef- fects of differential experience (light- vs. dark-rearing; environmental enrichment) on the rate and extent of expression of these genes could provide insight into the genetic mechanism of experiential modulation of brain organization, which has been theorized. The neural basis of learning and memory has become more approachable as a consequence of the advances in neural plasticity research out- lined here. The structural hypotheses arising from developmental and electrical stimulation experiments as well as biochemical hypotheses (e.g., the unique mRNAs) and electrophysiolog— ical hypotheses (e.g., field potential responsive- ness to afferent stimulation in the hippocampus) are just beginning to be examined in real-life situations. The basic conservatism of evolution suggests that developmental information storage mechanisms may be retained to subserve the memory process in adults; a few experiments support this view. Study of the modulation of the effects of ex- perience on the brain by hormones, neuropep- tides, and nonspecific neurotransmitters has assumed renewed importance with demonstra-— tions of (i) sex differences in the response of the brain to experience, damage, and functionally 1S acting electrical stimulation; (ii) involvement of catecholamines, sleep, and nonspecific arousal systems in regulating the effects of experience upon the brain; and (iii) continuing discovery of new neuropeptides with apparent modulatory roles in the adult nervous system. Studies of the first two classes of modulation have barely scratched the surface, and the results of more intensive investigation will almost certainly bear upon interpretation of the major clinical syn-— dromes (e.g., sex differences in developmental disorders and mental illness; interaction of dys— functioning catecholamine systems involved in mental illness with experience of therapy). It is imperative to understand the roles of neuropep-— tides in regulating intrinsic and experience- dependent aspects of brain development and the long-term effects of experience upon these systems. The development of appropriate animal models in this area offers potential insights into the mechanisms of early male vulnerability to brain disorder and postpubertal female sensi- tivity to schizophrenia and depression. Recommendations for the Future Explicit examples of research needs and oppor-— tunities are replete throughout this chapter. Summarized below are those of the panels' major recommendations that deal specifically with neural development and functional organization. @ Determination of the role(s) of trophic (survival) factors, neurite growth factors, cell surface and recognition molecules, hormones, and cell-cell interactions in neu— rogenesis and plasticity e Elucidation of the response of the brain to perinatal and adult insult in terms of both the initial injury and the capacity for re— generation and/or reorganization of the CNS @ Development of techniques of transplan- tation of tissue into the nervous system as an approach to replacing neuronal popula-— tions and facilitating neuronal regeneration @ Studies on neuronal aging, cell maintenance, and cell death in the aging organism e Determination of the precise changes that are associated with senescence, whether these changes are intrinsically or extrin-— sically generated, and how they may be modified @ Development of immunohistochemical, mi- crobiochemical, and functional fieuroana— tomical methods to unravel neural circuitry in the normal and diseased human brain ''Preparation of anatomical tools based on recombinant DNA technology as well as traditional biochemical methods Continuation of the cloning of genes coding for receptors, second messengers, trans-— mitter enzymes, neuropeptides, ion chan- nels, and characterization of brain specific genes Application of functional mapping methods to identify the specific brain loci related to mental processes and to study the develop— ment and organization of the neural cir-— cuitry underlying cognition, motivation, and affect Development and application of chronic in 16 situ techniques for measuring electrical and chemical activity of identified neurons and pathways in conscious behaving animals Development of quantitative morphological methods and computer technology to estab- lish a comprehensive textual and graphics data base of the anatomical and chemical organization of the brain Clarification of the relationship of circadian and ultradian rhythms of the sleep-wake cycle, hormone levels, body temperature, nutrients, etc., to neuropsychiatric diseases Studies of the influences of diet on behavior and the extent to which such influences are important in preventing or treating illness ''CHAPTER II NEUROTRANSMISSION: BIOLOGICAL AND PHARMACOLOGICAL ASPECTS Nerve cells communicate with one another primarily by means of the release of chemicals from the ending of one neuron onto specialized areas, or receptors, of another. Chemical neu- rotransmission is a common feature of all nerve cell networks in the animal kingdom. Through-— out, chemical neurotransmission has been found to entail a number of discrete molecular events—biosynthesis, processing, storage, and release of the neuroregulators; molecular recog- nition by receptors; transduction of the recogni- tion signal into biological events; inactivation; and metabolism. Indeed, the conservation of chemical neurotransmission in nature has per- mitted scientists to utilize information obtained from experiments on simple animals to generate hypotheses to explain aspects of the intricate human central nervous system (CNS). Other molecules also play critical roles in brain function. The electrical activity of the nervous system depends directly on certain pro-- teins residing in neuronal surface membranes known as channels; these channels control the flow of ions into and out of neurons. Contacts between neurons, upon which brain function de— pends, develop and are maintained through spe-— cific molecular interactions involving neuronal surface proteins and special factors, known as trophons, sent from one cell to the next. Even the shape of neurons is an expression of infor-— mation contained in brain deoxyribonucleic acid (DNA), and it has been speculated that learning and memory involve changes in gene expression. The fact that specific molecular properties are at the core of brain structure and function has been known for a long time. What is dif- ferent now is the availability of sensitive quan— titative techniques for studying these properties. New electrical recording methods permit the study of individual channel molecules with un- paralleled resolution. Recently perfected meth- ods for identifying, purifying, and determining the structure of brain proteins have become available. Techniques for isolating the DNA for individual brain proteins are now common and will yield not only details of protein structure but the organizational principles of brain mo- lecular architecture as well. The processes of neurotransmission and the 17 neuroregulators mediating these processes un-— derlie normal brain function and thereby sub-— serve the elaboration of mental activity. Many aspects of disturbed mental function--including disturbances in mood, thought, and behavior—can result from perturbations in the chain of neuro— transmission. For example, schizophrenia is thought to be related to abnormalities in dopa— mine (DA) function in one part of the brain, and Parkinson's disease to disordered DA function in yet another part of the brain. Alzheimer's dis— ease and other dementias have been linked to a loss of neurons utilizing the transmitter ace— tylcholine (ACh), and anxiety and stress appear to involve neuronal networks that use the transmitter gamma-—aminobutyric acid (GABA). The processes that occur at the junctions be- tween nerve cells, the synapses, are a focus of research, because clinically useful treatments have been shown to effect changes at this site. Hence, improved understanding of basic bio- logical and pharmacological aspects of neuro- transmission should lead to markedly better diagnostic tools and treatments for psychiatric disorders. Knowledge of the molecular principles under— lying behavior thus has important practical im— plications. For example, the fact that drugs can exert such specific effects as reduction of pain or anxiety is an expression of a molecular order in the nervous system that relates unique prop— erties of molecules to discrete states of self— awareness. To understand a drug's action, one must identify its target molecules and elucidate the role of these targets in neuronal structure and function. At the same time, insights can be achieved into the relations between these mo— lecular properties and behavior. One may use simple model systems, inverte— brates and lower vertebrates, to study the mo— lecular properties of the nervous system. In addition, new techniques permit the extension of these approaches directly to humans. This is true because: (i) Many important neuroregulator mol- ecules are present in peripheral tissue (blood cells or skin, for example) and can be studied there. (ii) Some molecules related to brain function appear in blood serum and spinal fluid. (iii) Metabolic and functional activity of the ''brain can be studied noninvasively. (iv) Samples of human neurons obtained at surgery or post mortem can be directly studied. Thus, new in- formation about the nervous system from a wide range of organisms and tissues may ultimately form the basis for diagnosis and treatment of disordered brain functioning of humans. Transmitter Identity Progress in identification of putative trans-— mitter compounds has been enormous during the past decade. The number of transmitter candi- dates has grown from a few to several dozen, and the list is not yet complete. Currently, both the evidence that these substances actually function as transmitters and the information on their physiological roles are rudimentary. De- scriptions of some of the better characterized transmitters follow. Acetylcholine. ACh is the classical trans- mitter substance released by motor neurons at vertebrate neuromuscular junctions. Much evi- dence supports a transmitter role for ACh in the vertebrate CNS as well. Of particular interest is the suggestion that two clusters of large basal forebrain cholinergic neurons (magnocellular neurons) project to the cortex degenerate and lead to the cognitive and behavioral abnormal-— ities of Alzheimer's disease. Amino Acids. Glutamic acid, aspartic acid, GABA, and glycine are believed to serve as major transmitter substances in the vertebrate CNS. The most compelling evidence for a transmitter role for two of these, GABA and glutamate, came originally from studies with invertebrate neuromuscular preparations. Much of the evidence in the vertebrate CNS comes from studies demonstrating a calcium—dependent release of these substances from electrically stimulated tissue slices and other CNS preparations. Other supportive studies for amino acid neu-— rotransmitter roles come from studies on the uptake and localization of radioactive amino acids, the characterization of receptors, and other physiological and pharmacological exami-— nations of the function of locally applied or neurally released substances. The case for GABA as a transmitter is further supported by micro— chemical assays for the enzymes and substrates of GABA metabolism, and by immunocytochemi- cal demonstrations at the light and electron mi-— croscopic level of the localization of glutamic acid decarboxylase, the enzyme that forms GABA. The development of specific markers for neurons that are likely to use the other amino acids as transmitters would be very important additions to the armamentarium for transmitter identification in the vertebrate CNS. 18 Amines. Serotonin (S-hydroxytryptamine or 5-HT), DA, norepinephrine (NE), epinephrine (E), octopamine, and histamine (H) all are believed to serve as neurotransmitters in the vertebrate nervous system. The most compelling evidence exists for the first three: Fluorescence cyto- chemical methods reveal that axonal projections and axonal arbors emerging from relatively small clusters of neurons within the brain stem ramify widely to many areas of the CNS, including the spinal cord. Amines have been implicated in many impor- tant processes in the CNS, but there is little understanding of their precise role in sleep, af- fective disorders, schizophrenia, memory for-— mation, and regulation of motor activity. In the latter case, however, DA neurons have been clearly implicated in Parkinson's disease, a dis— order characterized in part by tremors and im- pairment in the initiation of movement. In Parkinson's disease, DA neurons of the sub- stantia nigra that project to the basal ganglia degenerate. Replacement therapy, by oral in- gestion of dihydroxyphenylalanine (DOPA) (the biosynthetic precursor of the amine), has been a highly effective treatment for this disease in its early stages. Peptides. The concept that neurons are able to synthesize and secrete small peptides is not new. The ability of hypothalamic neurons to manu- facture oxytocin and vasopressin (AVP) and to release these hormones from nerve terminals in the pituitary has been known for over 20 years and established the identity of so-called pep— tidergic neurons. Current interest in the role of peptides in neuronal communication is a result of the availability of increasingly sophisticated methods for detecting, isolating, and mapping the distribution of peptide molecules within the nervous system. Twenty to 30 small peptides have been characterized and identified immu- nologically within sensory and autonomic neurons and in several CNS pathways. The rate at which new peptides are being discovered shows no sign of slowing. In the future, it would not be sur- prising to find several hundred different peptides identified within discrete subsets of neurons subserving different functions in the transfer of information in the nervous system. The molecular size of peptides found in ver- tebrate CNS neurons ranges from small dipep- tides like carnosine (found in primary olfactory neurons) to much larger substances. The list in— cludes substance P, the opioid family of peptides (met- and leu—enkephalin, B-—endorphin, dynor- phin, others), vasoactive intestinal peptide (VIP), cholecystokinin (CCK), neurotensin, somatosta-— tin, luteinizing hormone releasing hormone (LHRH), thyrotropin releasing hormone (TRH), adrenocorticotropic hormone (ACTH), calci- tonin, and many others. Most of the peptides found in nonneural tissues of the body have also ''been found within specific subsets of neurons in the vertebrate CNS. In some cases, multiple peptides are found associated with single neu-— rons; in other cases, peptides are colocalized and coreleased with small molecule classical neuro— transmitters. In still other cases, a single peptide may be the sole substance released by neurons. Identification and structural characterization of peptides is the first step toward understanding their biological function. Currently, the field is influenced by three approaches to the isolation of new peptides. The most classical approach examines the chemical basis of a particular physiological response by attempting to isolate the molecules that mediate the response. This approach has been particularly productive in the isolation of peptide hormones and releasing factors, leading, for example, to the charac-— terization of substance P, neurotensin, somato-— statin, and corticotropin releasing factor. With increasingly sensitive microsequencing tech-— niques, the introduction of fast atom bombard- ment—mass spectrometry and high-resolution chromatographic separation methods, peptide isolation and sequencing will be feasible with progressively smaller amounts of starting mate- rial and is therefore likely to acquire increasing specificity and importance. A primary advantage of this approach is that at least one of the phys-— iological actions of the peptide is established as soon as isolation and identification are complete. A second approach searches for endogenous compounds for receptors previously identified by pharmacological agents. An example of this ap— proach is the discovery of endorphin following the initial description of the binding of morphine to specific neuronal receptors. In a like manner, research is currently focusing on an endogenous neurotransmitter for the benzodiazepine receptor. The third and newest approach isolates the genes encoding neuroactive peptides and then obtains the entire amino acid sequence of the primary translation product containing the pep- tide and its precursors. This approach has been crucial in identifying multigene families coding for structurally and functionally related opioid peptide hormones, the hormones involved in egg-— laying behavior in Aplysia (see chapter III), and calcitonin gene—derived peptides. It has recently become possible to use the same cloning tech— nology to isolate peptide genes in specific se— cretory neurons without prior knowledge of the peptide hormone. Both immunological and chemical methods are employed in this field. For example, the ability to generate monoclonal antibodies directed against peptides and other transmitter candi- dates has meant that the distribution of these substances can be assessed quantitatively by radioimmunoassay and qualitatively by immuno-— cytochemical techniques. Such studies provide the anatomical detail necessary to enhance un-— 19 derstanding of the physiologic roles of a peptide and to suggest new functions for it. Radiolabeled complementary DNA (cDNA) probes can be hy- bridized to fixed tissue, thereby localizing cells synthesizing a given precursor (in situ hybrid— ization). This technique demonstrates that the peptide of interest is actually synthesized in a particular cell, rather than being accumulated through uptake or binding. Transmitter Synthesis The biochemical mechanisms responsible for the synthesis of neurotransmitter molecules are key processes through which information trans-— fer between neurons can be regulated. Study of sites of biosynthesis has led to the design of psychoactive drugs that will affect discrete subsets of neurons. In addition, it is now clear that neuronal impulse frequency plays an im- portant role in the regulation of enzyme activity by triggering intracellular events that eventually lead to the activation of the synthesizing en- zyme (e.g., phosphorylation of tyrosine hydrox- ylase). Another mechanistn by which regulation of transmitter synthesis occurs involves the re- * lease of diffusible factors by target neurons or effector cells that can alter enzyme activity (e.g., choline acetyltransferase). The recent dramatic advances in the rapid and accurate sequencing of cDNA have led to an explosion of information on the structure of the biosynthetic precursors of many bioactive pep-— tides. This has supplemented information from previous protein/peptide work. One startling finding has been the discovery of alternate splicing in messenger ribonucleic acid (mRNA) processing, which results in two or more families of related peptides from the same gene. Another important finding has been the demonstration that the same gene is often expressed in two or more tissues. Those tissues then sometimes produce distinct sets of peptide products with distinct biological activities. For other peptides, one sees a family of genes coding for structur- ally related peptides. Studies using recombinant DNA _ techniques have shown that small neuropeptides are syn- thesized in the form of large polypeptide pre-— cursors. Determination of amino acid sequences of the precursors culminated in the discovery of an important new class of proteins called poly- proteins (or polyfunctional proteins). They are precursors of more than one peptide—indeed, as many as eight different bioactive peptides, which have to be excised from the precursors to become active. These peptides then may act in concert to coordinate behavior. Thus far, structural analysis has revealed some of the arrangements of a few of these genes (including those for the three opioid peptide precursors, pro—-opiomelanocortin (POMC), pro- ''enkephalin, and pro-dynorphin) and indicated some intriguing evolutionary relationships. Re- petitive homologous elements within the genes coding for several neuroendocrine peptide pre— cursors suggest that these genes have evolved by duplication and rearrangement events. Progress has also been made in identification of end-product peptides (posttranslational proc- essing). Since all precursors to bioactive peptides are, by definition, bigger than their progeny, more than one product peptide can and usually does result from proteolytic processing of a precursor. The gene structure lays the frame- work within which the cell has to work, but many different steps can then occur. Tissue differ-— ences in the processing of POMC between the anterior and intermediate lobes of the pituitary have been described, and interaction between POMC processing products in the elicitation of a single biological response has been noted in the enhancement of ACTH-stimulated steroido— genesis in the adrenal cortex. Proteolytic proc- essing frequently occurs at pairs of basic amino acid residues, but not all pairs of basic residues are cleaved and not all cleavages are at pairs of basic residues. A recent study on the processing of prosoma-— tostatins by fish pancreatic tissue shows that cleavage leading to somatostatin-14 occurs at a lysine residue in one precursor, but not in a second, closely related precursor that produces instead somatostatin-28. Furthermore, a cleav- age site in a precursor may be utilized in one tissue and yet not be used as a cleavage site ina second tissue producing the same precursor. Ex— amples of differential tissue-specific processing of neuropeptide precursors have been found in the enkephalin system. Much careful biochemical characterization of peptide products remains to be done. An important but technically difficult problem will be to design methods suited to asking which specific form of a peptide is pres- ent at a synapse of interest, as opposed to ask— ing what the dominant forms of the peptide are in extracts of a whole block of tissue. The posttranslational processing enzymes are expected to be a major site of regulation, en- abling different sets of end-product peptides to be produced from a single precursor. Important questions include whether there are only a few such enzymes performing similar steps in dif- ferent tissues, or whether there are many dis— tinct enzymes, possibly even several for each distinct precursor. Studies of biosynthetic proc-— essing enzymes are confused by tissue or cellular heterogeneity—-the cells making the peptide of current interest are buried among many other cells. In addition, the subcellular sites of proc- essing—endoplasmic reticulum, Golgi apparatus, secretory granules—are difficult to obtain in large enough quantities and in sufficiently high purity for unambiguous results. Given the structure of a given precursor and a 20 detailed knowledge of the chemistry of its sev— eral product peptides, the biosynthetic pathway expressed in a cell needs to be determined. Such a pathway is not yet known for most bioactive neuropeptides, and thus far the pathway for any one neural peptide is not known in the same exquisite detail as for the better known amine neurotransmitters or hormones, e.g., insulin, parathormone, ACTH, endorphin. The crucial steps in such a pathway will be those that differ between tissues that use the same precursor to produce distinct sets of peptides. The major problem in studying such pathways seems to be the establishment of tissue cultured neurons that accurately mimic events seen in the animal. Transmitter Storage and Release In recent times, two model systems of storage and release have allowed profoundly detailed biochemical studies, because they yield large amounts of pure material needed for biochemical analysis of storage and release. These model systems employ the Torpedo electoplax, a cho- linergic presynaptic tissue, and the mammalian adrenal medulla, an adenergic tissue with mul- tiple presynaptic properties. Other systems such as brain synaptosomes, sympathetic ganglia, and nerve cell cultures have provided much useful information but suffer from the problem of heterogeneity. Granules from splenic nerves have also provided information on maturation and structure of noradrenergic vesicles, com- plementing data obtained from adrenal medulla research. The limitation here is that biochemi- cally useful amounts of material are obtained only by heroic measures. Platelets, containing S-HT granules, have also been an interesting model system, although functionally and struc-— turally distinct from nerve terminals. The neu- rotransmitter status of H and insulin are now being considered, but the relevance of mast cells and islets of Langerhans, respectively, to spe- cific terminals in the brain remains unexplored. Some key issues with respect to storage and release are (i) the involvement of cytoskeletal elements in directed movement of synaptic vesicles, (ii) the regulation by calcium (Catt) of the interaction of vesicles with cytoskeletal elements or with plasma or other vesicle mem- branes, (iii) the energetics of secretion, (iv) the details of membrane interaction and fusion, and (v) fundamental questions of how vesicles are assembled and the entire process regulated. Tubulin may play a role in movement of ves— icles (neurotransmitter storage structures within the neuron). Tubulin monomers are highly en— riched in presynaptic terminals and bind to chromaffin granule microtubules, providing the compelling notion that microtubules are a “railroad track" for vesicles riding to the plasma membrane. Tubulin polymerizes to form micro— ''tubules by a Catt sensitive process, possibly mediated by calmodulin. The role of tubulin in presynaptic function remains intriguing but obscure. The commonly accepted assumption is that Cat+ somehow causes synaptic vesicles to bind to and fuse with the plasma membrane. How this happens in molecular biochemical terms is under exploration. It was initially thought that Catt usually acts by activating specific protein mes-— sengers. Calmodulin is one such messenger that has been suggested as a mediator of Catt ac-— tion during secretion. It has a binding constant for Cat+ of 10-6M, a possibly relevant con- centration range for neurotransmission. Anti- calmodulin drugs such as the antipsychotic phenothiazines block neurotransmission, although they are by no means specific. Anticalmodulin antibodies, introduced into cultured chromaffin cells, are reported to inhibit catecholamine secretion. There is clearly an energetics of secretion, for’ the process appears to require adenosine tri- phosphate (ATP). Indeed, fusion in certain lipid membrane and liposome model systems seems to require an osmotic gradient across the mem- branes. However, the nature and origins of the putative osmotic gradient in cells remain to be completely explained. The need for ATP also indicates other possible pathways, including phosphorylation of proteins or lipids by specific kinases, or transmethylation of proteins or lipids using S—adenosylmethionine. Labeled products have been detected during secretion, and, in some cases, cause-effect relationships have been inferred from these data. This is an area of profound importance and uncertainty. Questions of vesicle assembly are related to the biochemistry and ultrastructure of the cell. One critical, widely discussed idea is that after secretion there is rapid vesicle recycling, so that old vesicle membranes get refilled with trans— mitters for subsequent resecretion. The evidence for this process is largely ultrastructural and has been obtained by some investigators, but not all, from studies on the frog neuromuscular junction. Studies pertaining to vesicle assembly using mature, isolated cholinergic vesicles or chro- maffin granules have seemed to emphasize the biochemical complexity of assembly, as well as some similarities between the two systems, and may yield insights into vesicle recycling. In the case of the Torpedo cholinergic vesicle, ACh is synthesized from mitochondrial acetyl coenzyme-—A and cytoplasmic choline. The ACh is then inserted into the vesicle by means of a hydrogen ion (H+) pumping, bicarbonate ion (HCO3) stimulated ATPase. In the release process, the macromolecular vesicle contents may not be secreted, but intragranular ATP is and thus must be replenished. The coexistence of ATP in these vesicles presages the concept that many transmitters may share a granule. The 21 suggestion that mammalian synaptic vesicles may contain little or no ATP has also been made, but mammalian synaptosomes also cosecrete ATP with ACh. Such problems may be tied in with an unresolved conflict in the field over whether nonvesicular secretion of acetylcholine can indeed occur. Transport of catecholamines into chromaffin granules appears to be energized by a Ht pumping ATPase that drives the substrate through a 45K molecular weight transport mole-— cule. In chromaffin granule ghosts, net accumu-— lation occurs, while in intact granules, transport involves a 1:1 exchange for an endogenous cat- echolamine. Transport systems have also been described for ATP and Catt, two core com- ponents, but not for ascorbate, another core component. The granules also contain enkepha- lins, a proenkephalin macromolecular precursor, and proteases that can generate enkephalins from the precursor. A critical problem is the nature of the granule interior, for both recent electron microscopy and nuclear magnetic res— onance results indicate that the core is heterogeneous. Transmitter Coexistence One natural result of the cocreation of many peptides from one precursor is the colocalization of those peptides in granules and their subse- quent corelease. When examined carefully, equimolar corelease is usually found to occur. Since different processing enzymes presumably are involved in making different sets of end- product peptides, it is conceivable that a neuron could respond to changes in its cellular and mo- lecular environment by changing its pattern of biosynthetic processing enzymes. As a result, the biological actions of that neuron could change, all within the framework of a single neuron making one primary gene product. A variant on this idea would be that two synaptic output re— gions from one cell might contain and secrete distinct sets of end-product peptides. The demonstration that neuroactive sub- stances may exist multiply in single neurons was made in the mid-1970s by investigators engaged in cytochemical studies. Observations were made in both the CNS and the peripheral nervous sys- tem (PNS) of mammals as well as in neurons of certain invertebrates. Transmitters that coexist include ACh, the catecholamines, S-HT, GABA, and numerous peptides. Combinations that have shown up in recent experiments suggest that certain peptide combinations occur with one or another amine neurotransmitter. This has prompted the speculation that these combina-— tions may have specific significance. The fact that multiple neuroactive substances may be called into play by single neurons makes ''cellular and molecular interactions exceedingly complex. Circuits may be multiple, transient, and unique in time constants, chemistry, and connections. This calls to point the likelihood that future drug regimens will be even more complex. To treat disorders, one might consider regimens consisting of several neuroactive sub- stances. For example, in cases of Parkinson's disease refractory to the usual DOPA treat-— ments, one could speculate on the need to in-— troduce a compound simulating a peptide-related action. Such complicated regimens at first glance might suggest a confused pharmacopeia for the future; yet this is not so. Knowing the mechanisms of action of each of the related neuroactive substances alone, and knowing their action in concert, will provide the keys for ma- nipulating complex systems with drug analogs. Transmitter Inactivation The use of peptides as intercellular signals in the nervous system requires a mechanism for cessation of their action. Based on results from the study of neurotransmission mediated by low molecular weight amines, three possible path- ways of termination of neuropeptide actions at their receptors can be suggested—diffusion, reuptake, and enzymatic hydrolysis. At the present time, there is no compelling evidence to substantiate the primacy of any of these proc— esses, although inactivation by reuptake seems least likely. Since inactivation by diffusion of neurotransmitters away from receptors can still be argued to be of primary significance in mono- aminergic transmission, this possibility must be entertained in consideration of peptidergic neu- rotransmission. However, recent advances in the study of peptidase enzymes and in electro- physiological studies of neuropeptides suggest that enzymatic hydrolysis of peptides may be of physiological significance. A large body of de- scriptive evidence on the ability of neural ex- tracts to degrade neuropeptides has been gen-— erated over the past decade. Until recently, these findings have been difficult to interpret due to (i) the large number of peptides involved, (ii) the scarcity of information on purified en- zymes, and (iii) an inclination on the part of many investigators toward a _ high-—specificity model in which a specific peptidase should exist for each neuropeptide. Recent studies, however, suggest that a small group of membrane-associ- ated nonspecific metalloendopeptidases and metalloaminopeptidases may exist in synaptic membranes of neurons as well as in glial cells. Results from electrophysiological studies suggest a relatively prolonged time course of neuropep- tide action and slow dissociation from receptors. Kinetic analysis of the above-mentioned pep- tidases indicates that the enzymes exhibit 22 properties that are consistent with this time course. A key goal of the neuropharmacology approach is the manipulation of synaptic transmission processes. This can best be founded on knowl- edge of intimate molecular details of synaptic transmission. Given the importance of peptide— secreting neurons to the operation of the nervous system, it is essential to have detailed biochem- ical information about the neuropeptide families. As discussed, this includes knowledge of the processing enzymes, the end—product peptides, and the inactivating peptidases. Such detailed information will permit the development of pharmacological probes and therapeutic agents focused on peptide biosynthesis, receptor coup- ling, and inactivation. Transmitter Receptors Receptors are, by definition, the initial ele— ments in the responses of all cells, including nerve and effector cells, to specific stimuli. They are generally macromolecules, usually large proteins, with binding or recognition sites selective for a given chemical signal. Their af- finity for a chemical signal is normally appro- priate to the physiologic concentration of signal. Receptors may be bound to membranes or lo- cated within cell compartments. External, in-— ternal, cytoplasmic, and nuclear receptors have evolved to allow selectivity and specificity with respect to signal and response. Evidence of other mechanisms awaits discovery. How do receptors regulate cell function so as to modulate CNS structure and function? Some receptors are coupled with transmembrane pores (channels). Viewed simply, these may be switching devices allowing yes-no decisions at the level of the cell membrane. Some receptors are coupled to complex enzyme cascades that alter intra— and extracellular levels of metabo- lites. These may regulate metabolic state and/or activate specific cellular processes. Other re— ceptors are coupled via intranuclear acceptors with the genome itself. These again may be simple yes—no switches (make a product or don't) or complex modulators (prepare to make a prod- uct if signal x is followed by signal y). Much recent effort has been directed toward the identification of the many transmitter re- ceptors. Two groups of investigations have evolved--those in which receptors have been examined using subcellular preparations from brain or peripheral tissue (e.g., muscle endplate ACh receptor) and others in which receptors and their response have been studied in cell culture systems. The latter systems have facilitated experimentation by allowing researchers to control the environmental milieu, to isolate phenotypically identical cells, to examine mech- ''anisms regulating receptor expression, and, in some cases, to isolate genetic variants that have provided powerful tools in defining mechanisms of action of receptors. Acetylcholine Receptor. Among receptors, the nicotinic ACh receptor has a unique status. It is associated with the earliest studies of neuro-— transmission and with the birth of the concept of specific receptors. In recent times, it was the first neurotransmitter receptor to be charac-— terized biochemically. Research on the ACh receptor made early advances largely because of its accessibility in the vertebrate muscle end- plate and its abundance in the specialized elec- tric organ of electric fish. It is reasonable to expect that in the near future the function of this receptor will be explicable in terms of its molecular structure. Delineation of structural details is critical to understanding relationships among receptor subtypes and between different classes of re- ceptors. Thus far, only limited information is available regarding structural details of most neurotransmitter receptors. Few neurotrans- mitter receptors have been purified to homoge- neity, and fewer still have been obtained in sufficient quantity to allow detailed structural analysis. The ACh receptor from Torpedo elec-— tric tissue is the most completely characterized and appears to be more or less typical. The monomeric form has a molecular weight of about 250K and contains two ACh binding sites. It is composed of four types of polypeptide chains of apparent molecular weights: 39K (a), 48K (8), S8K (y), and 64K (8) in the stoichiometry a 2t.Bpys. The four types of chains show about 40—percent sequence homology. Nevertheless, they give entirely different peptide maps and are immu- nologically distinguishable. Comparing receptors from Torpedo, from Electrophorus electricus, and from mammalian muscle, investigators have found that the chains cross-react. Furthermore, the stoichiometry of the chains in receptor from Electrophorus electricus is the same as in re- ceptor from Torpedo californica. In these two species, well separated on an evolutionary scale, the receptor is an asymmetrical pentamer com- posed of two identical and three unique chains. The ACh receptor has three functional states: (1) the resting state, in which the affinity for agonists is relatively low and the channel is closed; (2) the active state, in which the trans-— mitter binding sites are occupied by agonist and the channel is open; and (3) the desensitized state, in which the affinity for agonists is rela— tively high, but binding of agonist does not lead to channel opening. On binding two molecules of agonist, the receptor undergoes rapid transitions (on the submillisecond time scale) between the resting (closed) state and the active (open) state. These fluctuations last a few milliseconds, until 23 the agonist dissociates. The receptor also un- dergoes an agonist-promoted transition, on the time scale of tenths of seconds to seconds, to the desensitized state. Agonists like ACh and competitive antagonists like d—tubocurarine and the curarimimetic snake toxins all act primarily at the ACh binding site. By contrast, the noncompetitive inhibitors act on a different site on the receptor. These com- pounds include local anesthetics, other aromatic tertiary amines, antimuscarinic agents, antiar-— rhythmic agents, psychotomimetic agents, an- tiviral agents, alkaloid toxins, detergents, and alcohols. On the basis of their effects on the kinetics of endplate single channel currents, it has been suggested that they bind within and block the open channel. There is evidence that some of these agents also bind to the receptor in the closed state and decrease the rate of opening of the channel. In addition, many of these agents promote the transition of the receptor to the high-affinity desensitized state. Despite the complexity of their actions, the noncompetitive inhibitors provide useful probes for functionally significant sites other than the ACh binding sites. Some of these sites may be within the cation—conducting channel of the receptor. Receptor Function and Type. Another function of the receptor that should be considered is its self-regulation of mobility, clustering, turnover rate, and localization. The control of these properties could be mediated by as yet uniden- tified sites on the receptor that are subject to covalent modification or bind either small ef- fectors or other proteins. There are differences in the net charge and in the antigenicity of junctional and extrajunctional receptors. The functional significance of these modifications and structural differences, however, has not yet been determined. The past decade has witnessed a rapid pro- liferation of radioligand—binding studies using brain and peripheral tissues. New radioligands, new receptors, tissue or regional distribution of receptors, effects of various regulators (e.g., ions, nucleotides, temperature, drugs) or of disease-related or lesion—induced changes in re- ceptor binding all have been described. Results from such studies help one to formulate and to answer the key questions of current research on receptors. Investigators using subcellular preparations have largely dominated the area of receptor re- search. These workers have emphasized studies of recognition properties of receptors, as iden- tified by binding of radio-labeled probes (ra— dioligands). Such preparations, however, gener- ally do not permit assays of receptor—mediated responses. Thus, it is problematic whether these binding studies detect true receptors with spe- cific physiological function or merely nonfunc— tional binding sites. To identify a receptor, one ''must have the appropriate probe and the ability to measure functional response. Recent identi- fication of corticotropin releasing factor (CRF) and putative CRF receptors, using radio-labeled CRF in brain and pituitary, demonstrates how availability of the correct probe is essential for identification of CNS receptors. Much evidence supports the notion that the nervous system has evolved so that a single neurotransmitter can provide multiple types of signaling to different cells or even the same cell through interaction with receptor subtypes. Ex- amples include but are not limited to: adrenergic receptors (a1, «2, 81, B 2); cholinergic receptors (muscarinic M;,M2,My and nicotinic, central and neuromuscular); 5-HT receptors (S;, Sz); DA receptors (Dj, D2); and opiate receptors (u, x, §, and o). In some cases, these receptor subtypes, which were originally defined by differences in structure-activity relationships of agonists and antagonists, have also been shown to be linked to distinct | second—messenger systems, e€.g., «1-adrenergic receptors, increased cytoplasmic Cat+; ag-adrenergic receptors, inhibition of adenylate cyclase activity and decreased cellular cyclic adenosine monophosphate (cAMP); and By- and B2-adrenergic receptors to stimulation of adenylate cyclase activity and increased cellular cAMP. Thus, a _ particular neuro- transmitter (E or NE, in the case of adrenergic receptors) may have multiple ways to trigger or modulate target cell response. It would appear that evolutionary modifications have occurred in certain receptor types so as to yield subtypes with differences in both recognition and response. Since most receptors are present in mem- branes of target cells at low concentrations, purification and characterization by classical biochemical techniques have been technically difficult, laborious, and costly. Newer methods already have proved useful in supplanting these classical techniques. These include high—per- formance liquid chromatography, immunoaffinity chromatography, monoclonal antibodies, and cloning of the genes that code for the receptors. The power of this latter technique is exemplified by the recent success in cloning the genes for, and in turn inferring the amino acid sequence of, the subunits of the nicotinic cholinergic recep- tor. The striking homology between the subunits suggests that they arose from a common ances— tor. This important conclusion was rapidly de- rived from structural analyses of the genes of the subunits, which provided information that was not obtainable from isolation and sequencing of the receptor proteins themselves. In terms of function, it is extremely important to understand how neuronal and other target cells regulate their complement of receptors. The recent literature describing alterations in CNS receptors has rapidly increased, notably including reports of agonist-mediated "down 24 regulation"; antagonist-mediated “up regula— tion"; heterologous regulation (by other neuro- transmitters, drugs, or hormones); and changes in receptors with ontogeny, aging, or physiologic state. Notwithstanding the volume of the reports on these phenomena, detailed understanding of the cellular and molecular events (e.g., induc— tion, deinduction, and altered receptor degrada- tion) that mediate these processes remains to be elucidated. Moreover, most published reports have emphasized changes in receptor number, yet these changes are often small (less than 30 percent). Changes in receptor affinity for agon— ists or in postreceptor events may be of greater importance than these small changes in receptor number in producing rapid alterations in target cell sensitivity, as well as in providing amplifi- cation in regulation of target cell response. Receptor Pharmacology. Brain receptor pharmacology made great strides in the 1970s when the introduction of tritium labeling in ligand-binding assays made accurate determi- nation of brain receptor number and kinetics possible. The method rapidly gained popularity, because it could be applied to the study of vir- tually all brain neurotransmitter and drug- binding sites. Ligand-binding assays demon- strated (i) that neuroleptics may effect clinical state by blocking DA receptors in the brain, (ii) that many so-called side effects of psychother— apeutic drugs are due to their interaction with transmitter receptors, and (iii) that both drugs and diseases can alter receptor recognition sites and thereby be of use as primary screens in drug development. The receptor assays have also been adapted for use in analytical procedures for measuring tissue concentrations of neurotrans— mitters and drugs. Ligand-binding assays have made it possible to study receptor regulation and the molecular and anatomical characteristics of receptor sites. For example, studies of the binding of benozodiaze— pines to brain tissue indicate molecular inter— action between the high-affinity binding site (receptor) for the benzodiazepines and a recep— tor in the brain for the major inhibitory trans— mitter GABA. Recent investigations at a mole- cular pharmacological level have also indicated that these two binding sites are associated with a third, the ionophore site. This site seems to be associated with channels that are involved in chloride conductance in neurons, and it has been shown that a large number of convulsant com-— pounds interact at this site. Although these molecular mechanisms suggest that the three-part structure is a primary site of action of the benzodiazepines, it still remains unclear whether there are other sites at which benzodiazepines act. In particular, recent in- vestigations have indicated that at least one other high-affinity benzodiazepine binding site exists in most mammalian tissues. This so—called ''peripheral type binding site does not correlate as well with the anxiolytic/sedative properties of the benzodiazepines. In addition, other inves— tigators have indicated that lower affinity binding sites for the benzodiazepines, perhaps associated with a calmodulin—activated protein kinase, may also be important in some of the anticonvulsant actions of the benzodiazepines. This site and/or the peripheral site may also be involved in some of the known antiproliferative effects of the benzodiazepines seen at fairly high concentrations with transformed cells in vitro. In this case, the benzodiazepines are able to block cell growth and induce cells to differentiate. Numerous reports have described changes in neurotransmitter receptors in mood disorders, schizophrenia, and other neuropsychiatric con-— ditions. As yet, however, few studies have an- alyzed receptor changes prospectively and in a way that separates the impact of the disease of interest from treatment for the disease. More— over, studies in humans have generally been limited to the analysis of peripheral blood cells, cells that may not accurately reflect receptor expression at key regions in the brain. New techniques, such as the use of short-lived radio-— chemicals, computerized image intensification, positron emission tomography (PET), and NMR imaging may prove well suited for studies of receptors in vivo and may help define which psychiatric conditions are CNS receptor diseases. Receptor modifications may be responsible for the symptoms of a variety of psychiatric dis— orders, as well as the therapeutic responses to psychopharmacological agents. It is necessary to examine the properties of receptor regulation and their relationship to psychiatric disease and psychotropic drug action. The following areas have the’ greatest potential for future development. @ Identification and characterization of non- transmitter receptor modulators (e.g., GABA modulin and steroid hormones) ® Studies of the signals that determine the rate of receptor synthesis and degradation and the manner in which turnover is altered by drugs, hormones, and disease @ Development of in vivo methodologies (e.g., PET, NMR) that can be applied to clinical situations for visualization of receptors in human subjects ®@ Determining which receptor—mediated event is associated with which receptor—mediated function ® More detailed receptor maps using more selective, higher resolution probes (e.g., antireceptor antibody monitored at the 25 electron microscopic level) ®@ Cloning the genes coding for receptors ® Investigation of trophic factors for receptor expression and for cell—cell communications Postreceptor Coupling Mechanisms. How is occupancy of receptors by neurotransmitters translated into target cell response? It is in- triguing that although at least 50 different types of neurotransmitter/neuromodulator receptor types and subtypes have been identified, a much more limited number of biochemical mechanisms have been described for triggering receptor- mediated events. This suggests that postreceptor coupling mechanisms have been highly conserved through evolution. One of the best understood of these coupling systems is the regulation of adenylate cyclase activity, which appears to require at least two distinct guanine nucleotide (GTP) binding pro- teins (G) for either stimulation (Gg) or inhibi- tion (Gj) of cAMP synthesis. Because many different types of receptors appear to share a common ability to interact with these binding proteins, such receptors likely share a common structural domain. Perhaps receptors will thus have common and variable regions in a manner analogous to immunoglobulin molecules. Simi- larly, it appears that the binding proteins, for which structural and functional details are just now becoming available, appear to share fea- tures in common with each other as well as with other GTP binding proteins. These related bind- ing proteins are involved in photoreception, protein synthesis and chain elongation, and mi- crotubular assembly. Examination of the molecular details of re— ceptor coupling to adenylate cyclase is useful, because a number of transmitters either activate or inhibit adenylate cyclase activity. In both situations, receptor occupancy is translated to changes in catalytic activity via a multisubunit GTP-binding protein. The GTP-binding protein mediating activation of adenylate cyclase (Gg) is a complex possessing a 45K and a 35K Mr subunit. GTP occupancy results in expression of adenylate cyclase activity by stabilizing asso- ciation of the 45K Mr subunit with the catalytic moiety (C) of adenylate cyclase. The GTP-—bind- ing protein involved in cyclase inhibition (Gj) possesses a 41K Mr GTP-binding subunit that is associated with a 35K Mr subunit that is struc- turally indistinguishable from the 35K Mr subunit associated with G,. The precise mechanism by which Gj; produces inhibition of adenylate cyclase is not yet known. Several experimental procedures for evalua- ting coupling at the level of the receptor binding protein have gained widespread use. These pro- cedures have been useful not only in demon- strating the partial reactions involved in hor- ''mone-induced activation or inhibition of cyclase but also in evaluating the possible loci where changes in function occur subsequent to receptor desensitization. Desensitization may be homolo- gous (loss of sensitivity to only the agent to which target tissue has been chronically exposed) or heterologous (loss of sensitivity to all agents that share the effector mechanism by which the desensitizing agonist elicits its effects). In one model system for homologous desensitization to an agent that stimulates adenylate cyclase, de- sensitization appears to result from sequestering the f-adrenergic receptor from the other components, G and C, of the adenylate cyclase system. In contrast, heterologous desensitization studies in another model system are accom— panied by a cAMP-dependent phosphorylation of the B-—adrenergic receptor. Another coupling mechanism that is common to many receptors is phosphorylation of endog- enous proteins by protein kinases. This can be accomplished through several processes. Much work shows that cellular cAMP (and cGMP) levels regulate the degrce of phosphorylation of target cell proteins through changes in the ac-— tivation of a cAMP (or cGMP)-dependent protein kinase. The identity of the cyclic nucleotide do- main, i.e., the proteins susceptible to this regu— lation in the CNS, is gradually being defined. Another phosphorylation mechanism common to several other types of neurotransmitter recep-— tors, which involves increases in cellular Catt, acts to trigger cells through the action of its intracellular receptor calmodulin, which in turn may directly regulate cellular enzymes or may activate a Catt-dependent protein kinase. A third recent example of receptor—-mediated phosphorylation is the autophosphorylation of tyrosine residues in receptors, especially those for peptide neurotransmitters and hormones. These receptors commonly show agonist—induced clustering on the cell surface and a subsequent endocytosis. Many further molecular details of these mechanisms, as well as those for another general means of receptor triggering—trans-— location of receptors into the nucleus and alter- ation in the rate of transcription—are under active study. Of future importance is determin— ing the way in which ion channel-linked recep- tors (discussed later) and second messenger-— linked receptors influence each other and regulate each other's response. Most of the present proof that membrane protein phosphorylation plays an important role in certain transmitter and other actions has come from studies in molluscan nerve cells, specifically Aplysia and Hermissenda. The basic idea is that cyclic nucleotide-dependent sub- strate specific protein phosphorylation might be the general mechanism for action of second messengers. Mollusks are advantageous because the cells are generally large, allowing pressure injection of the protein kinase catalytic 26 (PKC) subunit or inhibitor, and identifiable, so that a homogeneous population of test cells can be used that are involved in known behavioral or physiological functions. Using this approach, it has been shown that S-HT action on two dif- ferent types of cells was mimicked by injecting PKC or blocked by injecting the inhibitor. It has also been determined that a long-lasting (30- minute) program of pacemaker discharge in a homogeneous group of peptidergic neurons trig- gered by a few seconds of synaptic input is es— sentially mimicked by PKC injection. In addition, catecholamine membrane effects can be mim- icked on heart cells by PKC injection. In many of the studies, there is associated biochemical evidence for specific protein phosphorylation in these different systems. However, it is still an open question as to the relationship between the phosphorylated proteins and the specific en- hancement or shutdown of ion channels. There is renewed interest in studies on the functions of phosphoinositides in the CNS. Phosphoinositol turnover in the nervous system is affected by nearly every hormone that elevates cytosolic Catt which, in turn, increases phosphoinositide turnover. The phospholipase C type enzyme (phosphodiesterase) appears to be activated and breaks down phosphoinositides to give the appropriate inositol phosphates. Cur-— rently, emphasis is on diacylglycerol as a mes-— senger, since this compound activates a cAMP- independent protein kinase C in the presence of Cat+ and phosphatidylserine. Breakdown of phosphoinositides to diacylglycerol is accom- panied by increased synthesis of phosphatidic acid as well as phosphatidylinositol, which is responsible for the increased uptake of radio- labeled inorganic phosphate noted in brain synaptosomes and other cells. The antimanic drug lithium (Lit) decreases the concentration of myoinositol in the cerebral cortex of rats. This decrease is accompanied by a marked increase in the concentration of inositol-1—phosphate. These findings suggest that Lit may act by inhibiting the conversion of inositol-1-phosphate to inositol by a phospha- tase. However, it may also stimulate muscarinic cholinergic activation of phosphatidylinositol breakdown. In the presence of Lit, phenylephrine, H, S5-HT, substance P, vasopressin, neurotensin, CCK octapeptide, and ACh all stimulate phos- phoinositide breakdown in brain slices. These data suggest that phosphoinositide breakdown is a widespread response coupled to a variety of receptors in the brain, and that it is probably a result of receptor activation and accompanied by an elevation of cytosolic Catt. Previous studies were hampered by the fact that in most mammalian cells the breakdown of phosphoinositides results in the release of ino- sitol, which is promptly used to resynthesize ''phosphatidyl inositol. In the presence of Li*, this process is often markedly reduced due to inhibition of inositol-1—-phosphate phosphatase. Future studies in the presence of Lit may prove most rewarding in the CNS. However, it remains to be determined whether phosphoinositide breakdown and elevation of Ca** have a causal relationship. Other aspects of phospholipid metabolism that show great promise include phospholipid methyl- ation and the role of phosphatidylcholine as a source of the choline for ACh formation in the brain. Undoubtedly, other enzymatic processes will be discovered that are affected by neuro- transmitter interactions with plasma membrane receptors. It is possible that hormones and neurotrans-— mitters work by mechanisms other than regula— tion of intracellular Catt and cAMP, but no other mechanisms have been established. The fact that most neurotransmitters affect the in- tracellular levels of one or the other of these messengers suggests that much more work is needed on these aspects of neurobiology. Ion Channels Ion channels are integral membrane proteins directly responsible for the electrical activity of the nervous system by virtue of their regulation of the movement of ions across membranes. The first report that single channels could be isolated on the membranes of intact cells occurred in 1976. This was made possible by the development of the patch clamp technique. In this technique, a fire—polished glass micropipette with a tip size around one micron is positioned on the mem-— brane of a neuron, and gentle suction is applied. The seal that develops between the glass and membrane, when sufficiently high (i.e., gigohms), allows the monitoring of tiny currents (a few picoamps) that represent the opening of one or a few ion channels. There are major advantages of this approach versus whole cell voltage clamp studies. For ex— ample, in the latter method, the summated po- tassium currents recorded are difficult to in- terpret because of the presence of five or six different types of these channels. In contrast, the membrane patch clamp technique can isolate a single channel, so that the cell does not have to be poisoned with a cocktail of nonspecific blocking agents. The approach depends on naked cells; thus, it is usually used on cultured neurons. An important evolution of this technique is the excised patch in which a piece of membrane is plucked from the cell. One can obtain inside-out and outside—out excised patches. With an out- side—out patch, one can use the patch as a mi-— croscopic detector of spontaneous transmitter release, or one can expose either type of excised patch to pharmacological agents with precise 27 control of concentration. Two classes of gated channels are recog- nized—-voltage-gated (those channels that open and close in response to changes in the voltage difference across neuronal membranes) and ligand-gated (those channels that open when a ligand, usually a neurotransmitter, is bound). The voltage-gated channels are responsible for the nerve impulse and, in general, for coding of in- formation and its transmission within the nerv- ous system. Ligand—gated channels are involved in synaptic transmission and thus play a central role in cell-cell communications within the brain and in the integration of information by neurons. Recently, a few examples of transition species of channels have been found, i.e., channels that are both ligand- and voltage-gated. A variety of channel types are present in the nervous system. About 10 different species of voltage-gated channels have been distinguished, and over a dozen ligand-gated channels are cur-— rently known. Many more probably exist, with one or two types present for each different transmitter. A particular neuron makes use of only some fraction of the total number of chan- nel types available in the nervous system, and the abundance and distribution of channels differ greatly from neuron to neuron. At present, no rational classification system is available for the various channel types. Volt- age-gated channels are named according to the type of ion that best passes through, and ligand-gated channels according to the ligand that opens them. This classification scheme, however, does not necessarily reveal all of the biologically important similarities and differ— ences between channel types. Research perspectives on channels are con- veniently classified into three groups—gating, selectivity, and regulation. Gating refers to the molecular mechanisms by which channels open and close to control transmembrane ion flows. Conformational changes are thought to underlie gating, but the number of conformational states and the precise nature of the protein shape transitions are unknown. All channels are se— lective in the types of ionic flows they control; some channels allow calcium ions to traverse the membrane, others are permeable to sodium, etc. Although the physical basis for selectivity is understood, the precise molecular structure by which it is achieved is still unknown. Regulation refers to a large class of processes by which neurons vary the number, distribution, and properties of their channels. Some of these processes are metabolic and involve up—and- down regulation by varying rates of synthesis and destruction. Channels are not distributed uniformly over the neuron surface but rather are inserted and maintained in particular regions of the cell sur-— face. Each channel type has its own character-— istic distribution, and this can differ from one ''type of neuron to the next. In addition to these longer term regulatory processes, channels are also subject to rapid regulation through neuro— modulation. The examples of neuromodulation best studied so far involve alterations in channel numbers or functional properties through phos- phorylation by cyclic nucleotide-dependent pro- tein kinases. Little is known about ion channels in the CNS, but it is reasonable to apply by analogy knowl- edge about channel composition and structure gained from other systems. Ion channels such as the cation channels associated with the voltage sensitive sodium channel, the nicotinic ACh re- ceptor, and the proton channel of bacterio- rhodopsin are the best understood systems of this type. In the case of bacteriorhodopsin, a single polypeptide makes up the channel; in the case of the ACh receptor, a pentamer of homologous polypeptides contains the recognition site(s) for ACh and the cation channel. Such detail has not been resolved for the voltage-dependent sodium channel, which is known to be comprised of a very large polypeptide but may or may not con-— tain additional protein components. These sys— tems are the best examples of membrane—asso- ciated ion channels for which it may be possible to determine the structure(s) responsible for ion transport. It has been shown that membrane proteins do the job of transporting ions, but it remains to be determined what kinds of second- ary, tertiary, or quaternary arrangements of polypeptides effect transport of ions, voltage- dependent transformational transitions, or other membrane-related phenomena. An exciting new approach is the induction of ion channels by using mRNA. In one example, GABA receptor mRNA extracted from the optic tectum of chicks was injected into frog oocytes (egg cells). The cell synthesized and inserted into its own membrane GABA receptors. As judged by single channel recordings, the GABA receptors were functional. This approach has great poten- tial for detecting the function of mRNAs related to receptors and ion channels in different brain locations. By injecting the mRNA into frog oocytes, one has the potential of expressing functional receptor-ion channel complexes in a considerably more accessible physiological preparation. Our understanding of the topography of ion channels within the cell membrane is still in a formative stage. Immunocytochemistry at the level of electron microscopy may be used to de-— termine whether a system is transmembrane in location by using various antibody electron- dense conjugates. The folding of polypeptides through the membrane can be approached profitably only for systems for which the pri- mary structure has been characterized. The possibility of extending structural information further than this depends on electron diffraction methods or low-angle X-ray diffraction meth- 28 ods. These approaches depend on the clustering of membrane proteins. Until recently, there was little hope of achieving this for dispersed re— ceptors. A promising new method employs doping a phospholipid monolayer with a long-chain lipid containing a specific ligand for a specific pro- tein, so that it protrudes from the monolayer. The protein attaches to the membrane and can form a two-dimensional crystalline array. Such affinity—directed crystallization approaches may become very useful for diffraction studies of CNS proteins in the future. The cloning of genes responsible for genera— tion of ion—channel components in the nervous system is now a reality. These rapid and infor- mative approaches are amenable to complete primary structure determination, as in the case of the nicotinic ACh receptor, resulting in (i) the formulation of semistructural models for its in- teraction with the postsynaptic membrane, (ii) its expression in oocyte membranes and the po- tential for determination of the function of each subunit or combinations thereof, (iii) regulatory mechanisms for such receptors, and (iv) the possibility to use site-specific mutagenesis for investigation of critical regions of the molecule in terms of normal function as well as for studies of receptor biosynthesis and regulation. Such approaches will be a major area of endeavor in the study of CNS receptors. Our understanding of channels should increase enormously in the future because of the new techniques. In reference to gating, patch clamp recording provides good definition of the dif- ferent conformational states a channel can as- sume and of the transitions between these states. Thus, channel gating can be studied with high resolution. As more channel genes are cloned and sequenced, it will be possible to re- late physiologically defined conformational states with actual molecular structures. When cloned channel genes have been transferred to cell lines, site-directed mutagenesis will be possible as a method of studying structure- function relationships. Similar approaches should be productive also in studies of channel selec- tivity. Ion permeation can be defined quantita- tively with electrophysiological methods, and the structural basis elucidated by sequencing cloned genes and using this sequence information as the basis for appropriate biochemical approaches. The existence of multiple allelic forms of re— ceptors is an area that will be expanded as the techniques to approach these receptors as pro- teins expand. In this regard, techniques such as affinity or photo—affinity labeling of receptors yielding radioactive receptors will permit ap- plication of two-dimensional gel electophoresis and other techniques currently used for studying enzyme polymorphisms. If gene-—cloning ap- proaches are successful, even more variants can be defined. This is the case because not every mutation results in an electrophoretic variant ''(only changes in the net charge of the protein would result in these variants). Other changes— for instance, substitution of isoleucine for leu— cine-might affect function and would be picked up by a genetic screening technique. At a more general level, advances can be an- ticipated in two areas. Patch clamp recording makes it possible to distinguish channel types, which should lead to the discovery of new spe-— cies of channels. At the same time, cDNA clones for channels permit the definition of possible families that share structural and functional features. The combination of these two ap-— proaches promises to reveal a new order in the relation between channel types and thereby to increase understanding of function. A role for channels in altered brain function cannot be assessed until a better understanding of channel structure and function is achieved. Many psychopharmacological agents probably interfere with the ability of neurotransmitters to affect plasma membrane proteins. Fundamental studies on membranes should be of value in the development of more effective drugs. Especially important are studies on the solubilization and purification of neurotransmitter receptors, GTP regulatory proteins, adenylate cyclase, ion channels, and other proteins thought to be in- volved in signal transmission through the plasma membrane. Studies in systems simpler than those derived from mammalian brain may be equally valid and important. Genetics of Neurotransmission For each transmitter system, there must be mechanisms for synthesis of the transmitter molecule, for release of the transmitter mole- cule into the synaptic cleft, for receipt of the signal at the postsynaptic membrane, and for clearing the synaptic space of the transmitter so that the next impulse can be received by the postsynaptic membrane. Many of the operational steps in this system involve a polypeptide chain, and hence, a direct gene product. (In certain systems, the transmitter itself is also a poly- peptide chain.) Two major difficulties in sorting out polypeptides have been that the mixture of proteins isolated from the brain is exceedingly complex, and that many of the system compo- nents are bound to and are active only in con- junction with cell membranes, making them difficult to purify and assay as isolated compo- nents. The advent of recombinant DNA technol- ogy opens the door to a more complete evalua- tion of all of the important components of the neurotransmitter systems in the brain. This is so because the gene products can, in essence, be recreated through genetic engineering tech-— niques once the genes that encode them have been isolated and characterized. 29 Molecular Approaches. At present, two ap- proaches are being taken by molecular biologists to characterize significant elements of neuro- transmission in the brain. One approach is to identify elements whose behavior has already been characterized from a physical point of view. Examples of such molecules include sodium potassium ATPase, sodium channel proteins, and the ACh receptor. The cloning of these mole- cules should lead to the isolation of the genes for less abundant polypeptides that play an impor- tant role in the transmission of nervous impulses in the brain. The second approach involves the creation of libraries of either monoclonal anti- bodies or DNA copies of brain mRNA in order to assess systematically which proteins in the brain are localized to specific regions. The presump-— tion here is that specific function will be as— sociated with specific gene products. These approaches not only permit investigation of the structural elements of neurotransmission but also allow identification of the elements of particular functional cell units within the brain. The ability to identify gene products that are characteristic signatures of particular cell groups in the brain should lead to an under- standing of how the functional ability of a par- ticular area is encoded through the genetic program. The basic principle of cDNA cloning technol- ogy is that every polypeptide synthesized by a cell is coded for by an mRNA molecule present in that cell. Using the enzyme reverse tran- scriptase, synthetic copies of the mRNA mole- cules in a cell can be constructed. (These copies are termed cDNA.) The cDNA molecules can then be inserted into a DNA vector molecule which allows each cDNA molecule to be pre- cisely reproduced by introduction of the com- bined cDNA-vector molecule, into a host cell such as E. coli. A collection of cDNA molecules derived from a given cell type represents the complete array of polypeptides made by that cell type. The cDNA representing a particular gene product can be identified by appropriate selec— tion methods. The sequence of nucleotides in the cDNA can easily be determined. This sequence directly determines the structure of the poly- peptide chain. In this way, the composition of crucial elements in brain circuitry can be determined. The use of cDNA cloning techniques will eventually result in a complete catalog of poly- peptides that are significant in the control of the transmission of impulses in the human brain. This approach has already proved to be an important tool in identifying humoral factors that play im- portant roles in brain function. The advantages of the cDNA technology include the ability to determine the precise amino acid sequence of the polypeptide in question much more easily than by conventional techniques, and the ability to transfer the gene to other cell types in order ''to assess the biological properties of the poly-— peptide in systems more amenable to experi- mental manipulation than the human brain. The use of recombinant DNA techniques may also permit the analysis of the phenotype of in— dividual cells in much more detail than is now possible. For example, it is possible to produce cDNA libraries from mRNA isolated from in— dividual neurons in ganglia of lower organisms such as the lobster and Aplysia because of the large size of these neurons. One can then dif- ferentially screen these libraries to determine which of the recombinant clones are unique for a given cell type. The cDNA inserts can be se- quenced to determine the structures of the pro- teins or peptides they code for. One can then analyze the functions of the peptides by the techniques described. This approach might be extremely helpful in unraveling the highly com-— plex chemistry and physiology of the nervous system. The availability of cloned cDNA and genomic probes for neuropeptide precursors, receptor proteins, and ion—channel proteins will enable researchers to study the mechanisms of long-term plasticity, which occur at the transcriptional level or at the level of gene rearrangements. The probes can also be used to determine how levels of mRNA vary in response to administration of physiological agents and drugs. Effects of those agents often are more dramatic at the level of transcription than at the level of translation. Developmental Studies. Genetic approaches also have been uniquely useful in providing in- sights into mechanisms underlying development; what we have learned concerning the algorithms used in cell lineages in nematodes and develop- ment of segmentation in Drosophila could not have been derived from any other discipline. At the level of behavior, mutations affecting learning and memory in Drosophila appear to affect such molecules as _ phosphodiesterase, adenylate cyclase, or the enzymes and receptors involved in adrenergic transmission. Most likely, genetic studies such as these will lead to the discovery of previously unknown molecular en-— tities that are important for the development or function of the nervous system. The level of sophistication and power of ge- netic analysis is inversely proportional to the generation time of the organism. Drosophila, having a short generation time, are very ame- nable to genetic approaches. For instance, with the current state of the science in Drosophila research, as long as a gene can be mutated to give a recognizable phenotype, it can be cloned and subjected to detailed analysis of its expres— sion, regulation, etc. Alternatively, if one can localize the gene coding for an important neu- ronal element on polytene salivary chromosomes (this can be done, for instance, if that molecule 30 has been cloned in other species and shows enough homology with its counterpart in Droso- phila), one may induce mutations of that gene and study the functional or developmental con— sequences of the mutations. Moreover, it is now possible to mobilize a class of transposable elements (transposons) at will and isolate mutations caused by the inser— tion of the transposable element. This greatly facilitates cloning of genes in Drosophila. For example, a genetic locus that probably codes for structural components of a potassium channel has been mutated by the insertion of a trans— poson, so that the cloned transposon can be used to pull out DNA sequences from that genetic locus. Transposons and the machinery for transposition can also be used to introduce and integrate any cloned DNA sequences (which may be mutated in vitro) into the genome of a Dro- sophilia so as to generate stable, transformed stocks expressing the newly introduced gene (often with proper regulation of expression). Currently, features such as these are not matched in organisms other than bacteria, yeast, and possibly nematodes. Recombinant DNA technology is also useful in analysis of developmental events in the mam— malian CNS. Many analogies have been made between the diversity of the immune system and the complexity of the CNS. For example, the diversity of immunoglobulins derives from ele- gant DNA processing events. A similar diversity of cell surface molecules on nerve cells might account for their selective instructions in de- velopment and might be due to similar DNA processing mechanisms. Monoclonal antibodies could be used to detect these molecules, and recombinant DNA techniques applied to reveal processing events. Mental disorders of late onset may be due to developmental events that occur late in life, which could be detected by these molecular approaches. An essential aspect of the nervous system is communication among cells. This requires that the nervous system develop in an orderly way, so that the proper cells are brought in contact with one another and the proper neurotransmitter release and receptor mechanisms are synthesized in every cell. It is likely that these events re— quire regulation of the expression of genes in the developing nervous system. It is also likely that many psychiatric disorders are due to improper communication among nerve cells, and that this, in turn, is due to inadequate regulation by those genes coding for the components of neuro- transmitter mechanisms. Analysis of the regu- lation of gene expression in the nervous system requires tools to study the organization of the relevant genes and to detect their primary transcription products. Recombinant DNA technology is especially well suited to this task. Cloned genetic information coding for a par- ticular neurotransmitter receptor or neuro-— ''transmitter synthesizing enzyme can detect expression of the relevant genes during develop- ment. This, in turn, may facilitate understanding of how neurons regulate expression of their genes. If psychiatric disorders due to improper neuronal communication are inherited, these cloned sequences can potentially detect the relevant genes and thereby identify individuals at risk. Structural Determination. Traditional bio— chemical techniques are most useful when there is a rich source of the relevant molecules. Un- fortunately, many interesting molecules are present in very small quantities in the nervous system, making it difficult to isolate enough of them for studies to determine their structure. This is particularly true in cases in which the molecule is very large and large quantities are needed for structural analysis. In the case of a protein, however, the primary structure can be determined from the nucleotide sequence of its structural gene. This has two advantages: The gene can be isolated and grown in bacteria to provide an essentially limitless supply, and the nucleotide sequence of DNA is much more easily obtained than is the amino acid sequence of a protein. Even the most minor components of the nervous system can now be _ studied, their structure determined and sequenced, and spe-— cific probes prepared. For example, many neu~ rotransmitter receptors are pharmacologically closely related. If the primary structures of these receptors were available, one might find specific amino acid sequences that identify with specific pharmacological properties. Antibodies against these amino acid sequences or nucleic acid probes for the corresponding nucleic acid sequences would provide specific, reproducible ways of studying the different molecules. The fact that many proteins have common ancestors is reflected in their sequence homolo— gies. This homology can be exploited to detect related genes by low stringency hybridization techniques. This may identify genes whose gene products are such minor components of the nerv— ous system that they might not have been identified without first discovering a function. Alternatively, if genes concerned with a common process are linked in the genome, they might be identified by looking at sequences that flank known genes. By chance, a gene product that is genetically related to the acetylcholine receptor has been identified; it may be a component of the neuromuscular junction. In addition, new neuropeptides have also been identified by anal- ysis of the sequences adjacent to a known gene. Neuropeptides. The biological functions of new peptides are now being examined by generating antibodies against the putative peptide and studying its pattern of distribution via immuno- histochemical procedures. This approach may 31 prove that the peptide of interest exists natur- ally, but the pattern of distribution could no more than suggest a functional role. The intro— duction of the synthetic peptide or the antibody to that peptide into animals would be helpful in the search for a functional role. Gene transfer enables the creation of transgenic animals that overexpress the products of polyprotein genes. This type of experimentation would dramatically enhance the ability to determine the role of pep-— tides derived from polyproteins. The cDNA probes that have been developed for detecting species of mRNA that code for neuropeptides are also opening up new ap- proaches to the study of the development of the nervous and endocrine systems. Up to now, ex-— pression of a phenotype in a given cell or tissue during development has been limited to the study of the appearance of proteins or peptides. Tran- scription of some genes may occur much earlier during development than translation. With the highly sensitive cDNA hybridization methods now available, it should be possible to analyze very early stages of gene transcription, even when only a few copies of a particular mRNA are present. The hope is that in situ hybrid— ization with cDNA probes will allow scientists to determine, with great sensitivity, when in the course of development a particular gene is turned on in a particular neuron. This would make it feasible to follow cellular differen- tiation from a much earlier stage than is now possible. Recombinant DNA approaches, particularly gene transfer techniques, will greatly enhance the ability to define neuropeptide biosynthesis. Gene transfer experiments in the immediate future are likely to be of three types: (i) ex- pression of foreign genes in frog oocytes where assembly of ACh receptor subunits has been shown to be feasible, (ii) transfection or trans-— formation of cell lines by viruses or plasmids containing the genes of interest, and (iii) trans- fer of foreign genes into mouse embryos to make transgenic mice. Such experimentation could be used to identify the specific DNA sequences in- volved in transcriptional control and may also play a role in characterizing various aspects of the proteolytic processing reactions. Here, too, site-specific mutagenesis (altering dibasic amino acid cleavage recognition sites, for example) followed by gene transfer might be employed. Currently, an obstacle to the cloning of receptor genes is the lack of methods for screening transformed cells for the presence of receptor proteins. Drug Development. Recombinant DNA tech- niques also have significant potential for the development and testing of new families of drugs. A major problem in drug development is the lack of appropriate in vitro systems to directly test the biochemical interaction be- ''tween a drug and its receptor or other polypep-— tide with which it interacts. Through the use of genetic engineering techniques, this problem is being circumvented. The potential to develop improved test systems is based on the principle that a gene of interest can be inserted into a DNA molecule that contains signals for the transcription and translation of the inserted gene, as well as signals that allow the synthetic DNA molecule to replicate successfully in the cell into that it is introduced. The utility of such systems in their simplest form is to provide large quantities of a scarce macromolecule which is the potential target of a drug under develop-— ment. This capability would permit the direct biochemical analysis of drug—target interactions. At a more sophisticated level, the ability to introduce a series of active genes into a cell type of choice could result in the engineering of a synthetic target cell that could act as a direct analog of a target cell population in the brain. Mental Iliness. Until recently, hypotheses concerning molecular abnormalities underlying mental illness were difficult to test, because relevant molecules (e.g., receptors, channels) could not be readily purified. Progress has also been hampered by the lack of adequate high- affinity toxins or antibodies, and because the resolution of biophysical studies on central neurons is less than optimal. The new approaches discussed here make it possible to study ion channels at the molecular level by isolating the genes or messenger RNAs coding for them. Thus, conceivably, one will be able to compare the structure and expression of these genes in patients and in normal subjects. Mental illnesses like the affective disorders and schizophrenia appear to have a genetic base; these new tech- nologies may permit the genetic basis for such illnesses to be identified at the molecular level. If so, the field may progress to a stage similar to that of hemoglobin research concerning heritable diseases, where it is now possible to diagnose sickle cell anemia prenatally by analyzing the 32 distribution of restriction enzyme recognition sites in the globin genes. Recommendations for the Future Explicit examples of research needs and op-— portunities are replete throughout this chapter. Summarized below are those of the panels' major recommendations that deal specifically with biological and pharmacological aspects of neu- rotransmission. © Identification, characterization, and elu- cidation of the molecular basis of the function of novel neurotransmitters/ neuromodulators @ Clarification of the molecular mechanisms involved in processing and metabolizing neuropeptides e® Elucidation of the cellular and molecular mechanisms governing neurotransmitter/ neuromodulator storage, release, and metabolism ® Identification, localization, and functional and molecular characterization of receptors and receptor subtypes ® Elucidation of the mechanisms by which receptor-mediated signals are translated into biological events e Analysis of neurotransmitter/neuromodu-— lator mechanisms underlying behavior e@ Elucidation of the effects of psychophar- macological agents on all aspects of neu- rotransmission/neuromodulation e Elucidation of the relationship between neurotransmitter/neuromodulator biology and neuropsychiatric disease ''CHAPTER Ii HORMONES AND NEUROPEPTIDES IN THE CONTROL OF BEHAVIOR Hormonal influence on central nervous system (CNS) function, with an emphasis on relating CNS function to behavior, is the focus of many new and exciting studies. The hypothalamus is a known integrating center for such functions as hormone secretion, body temperature regulation, food and water balance, cardiovascular regula-— tion, and aggressive and sexual behavior. Much recent work has extended our understanding of the molecular, cellular, intercellular, and be- havioral aspects of homeostasis. The ultimate goal of these investigations is to understand the human brain and the possible role of environ- mental and humoral factors in behavioral disease states. Recently, several peptides have been de— scribed within the CNS. Some of these are pep— tides previously described within the pituitary and gastrointestinal tract; others are peptides previously characterized in invertebrates. Much of this information has been obtained within the past decade. It is now becoming evident that many of these peptides are synthesized within the brain, and their distribution patterns have been established by immunocytochemical meth-— ods. Recombinant deoxyribonucleic acid (DNA) technology not only has provided evidence of new peptide products in the CNS but has char- acterized precursor products within brain and, in some instances, provided definitive proof of synthesis. Information is still needed with regard to the processing of these peptides in brain and their regulation in various terminal fields. The function of many of these peptides is still un- known, although they appear to be involved in all of the major homeostatic systems of the body. Hormonal Effects A review of recent conceptual and methodo- logical developments is useful in appreciating how questions in neuroendocrinology are cur-— rently being addressed. In the early 1960s, small crystals of steroid hormones implanted in spe- cific areas of the CNS were found to trigger mating behavior changes in animals with very low levels of circulating hormones. Stimulated in part by the discovery that target tissues for steroid hormones contain cells that retain those hormones in a specific manner, work with cell fractionation of brain regions and steroid hor-— mone autoradiography of the CNS were system- 33 atically performed. It was subsequently found that tritiated estradiol was retained for hours in homogenates of hypothalamic tissue at high levels but not in control tissues such as the cer-— ebral cortex. Autoradiographic studies showed essentially the same thing by using counts of silver grains over the nuclei of cells in hypo- thalamic regions at different times following administration of tritiated estrogenic or andro- genic hormones. In the 1970s, work was done correlating hor- mone receptor properties with behavior and en- docrine changes. One question concerned the metabolism of the isotopically labeled hormone administered: Is it metabolized before it reaches the nervous system as a possible requirement for eventual retention by nerve cells, or after it reaches the nervous system as a consequence of hormone action? Progesterone has been a dif- ficult subject for study in this regard because of its position in the route of synthesis of a wide variety of steroid hormones. The clearest sets of conclusions have come from work on the metab-— olism of androgenic hormones. In some andro— genic hormone target tissues, reduction at a particular position is required for the entry of the hormone into the cell nucleus and for even— tual hormone action. In other cells, aromatiza— tion of the A ring to make an estrogen is re— quired for hormone action. There are also a number of instances in which the initially ad- ministered hormone is active in its native form. It is useful to note the differences between two major experimental paradigms, the endo— crinological and the correlational, each only partially useful in testing the role of specific neurochemical systems in definable behavioral phenomena. In the endocrinological paradigm, in order to assess the function of a brain compo-— nent (or endocrine organ), scientists focus on what happens or fails to happen when that cir- cuit is removed or activated. The adaptation of this paradigm to the functional analysis of neu— roanatomical structures, especially within the context of behavioral research, requires the as— sumption that a given behavior is a unitary phe- nomenon, and that only one system, clearly amenable to a lesion, is required for its suffi- cient execution. In the correlational paradigm, alterations in the basal activity of a system with a concomitant change in behavior are used to infer the participation of the system in that ''behavior. Invertebrates. Invertebrates offer many useful systems for the study of the interaction of hormones and behavior; in many invertebrate species, specific neurons have been identified that can be reproducibly isolated in successive experimental animals. Also, many invertebrate neurons are quite large and thereby allow bio- chemical and physiological studies on single cells. Many of the behavioral changes in such organisms are simple enough that one can po- tentially work out the entire circuit underlying the behavior. Such systems can be used as mod- els for phenomena seen in more complex nervous systems. Virtually all of the invertebrate work has been carried out in mollusks and arthropods. In the marine mollusk Aplysia californica, egg-laying as a behavior pattern has been extensively studied. The pattern consists of a 2- to 3-hour program of stereotyped behavior that can be evoked by injection of the secretions from a paired cluster of neurosecretory cells, the bag cells. Research with arthropods has shown ju- venile hormone affecting a variety of behaviors, e.g., those associated with maturation of female sexual responses, aggression, feeding, and domi- nance hierarchies. However, essentially nothing is known of the underlying neural mechanisms. Another steroid hormone, ecdysone, releases premetamorphic behaviors in larval insects, such as the tobacco hornworm (Manduca sexta), where treatment with ecdysteroids at the end of the last larval stage induces a daylong locomotor activity called the wandering behavior. Both peptide and steroid hormones play a role in the activation of invertebrate behaviors. Little is known about the structure of certain of these substances, e.g., eclosion hormone has been only partially characterized. A number of other peptide hormones are being studied in mollusks. In Aplysia, two atrial gland peptides have been isolated and sequenced. These pep— tides also release egg-laying behavior when in- jected into Aplysia, apparently through the stimulation of egg-laying hormone (ELH) secre-— tion. They appear to represent an endocrine link between the normal stimulus of egg-laying (i.e., copulation) and the subsequent activation of the bag cells. There is growing interest in examining inver-— tebrates for the presence of vertebrate peptides and seeking to identify a physiological role. One of the best documented is the action of vaso-— pressin (AVP) in inducing bursting activity in cell 11 in the CNS of the snail Otala. AVP appears to mimic a peptide that is endogenous to the CNS of the snail. Many studies in this area have utilized antibodies for specific vertebrate hor- mones. Using either radioimmunoassay or im- munocytochemistry, various researchers have reported a wide variety of antigenically similar 34 peptides in arthropods and mollusks. The next logical step is the purification and sequencing of those substances that have been identified. The best progress to date has been for a hormone isolated from the brain of flies. This pancreatic polypeptide-like material has not yet been se- quenced, but its amino acid composition is al- most identical to that seen for the mammalian pancreatic polypeptides. Genomic Regulation. Uncovering the manner in which steroids regulate gene expression in brain cells has been most difficult. The standards here derive from nonneural tissues such as chick oviduct (avidin) and frog liver (vitellogenin), in which steroids induce the expression of new messenger ribonucleic acids (mRNAs), and the protein produced is a considerable portion of the cell's output. Brain is a heterogeneous tissue which may contain many different sorts of ste— roid target cells. It has not yet been determined how new proteins crucial for the behavioral roles of neurons are produced or regulated through steroid interaction. One approach might be to examine gene activation in response to steroid challenge by looking at in situ hybridization of complementary deoxyribonucleic acid (cDNA) probes. cDNA libraries are being collected for a number of species; such a technique could serve to discriminate powerfully among target neurons. The revolutionary advances of molecular bio-— logical technology hold great promise across a wide range of biobehavioral research. It is now possible to inject molecules into cells or to transfer them into cells by vectors such as the SV40 virus for DNA and liposomes for enzymes. These techniques make it possible to add foreign enzymes or DNA to perturb cells in a known manner and to introduce mutated genetic ma-— terial isolated from one cell into a normal cell. By these means, highly specific changes can be made in various stages of the cell cycle. In ad- dition, monoclonal antibodies can be generated to constituents within the cell or to components of the cell surface. Such antibodies have already been used for cytochemical mapping and detec— tion of small changes in cell surface proteins and carbohydrates. These new technologies have made it feasible to study gene regulation in a complex system such as the brain. Thus, it is now possible to de- velop libraries containing individual copies of all of the mRNAs expressed in a particular neural tissue. By creating cDNA libraries from specific tissues in different behavioral states, it becomes possible to analyze the differences in the tissues at the level of gene expression. Identification of genes whose expression is altered is accom-— plished by a variety of techniques, including DNA sequence and computer analysis, hybridi- zation with known DNAs, and use of a positive hybridization/translation system where the DNA sequence is used to isolate its complementary ''mRNA, which is then translated into protein. As more specific DNA probes are isolated (currently at the rate of 10 new cDNA/genes per month), they expand the repertoire that can be used to analyze specific changes in gene expression resulting from altered behavioral states. For example, the cDNAs that code for the catecho- lamine biosynthetic enzymes have now been cloned in bacteria. Thus, it is now possible to measure levels of tyrosine hydroxylase mRNA in the preoptic area of the hypothalamus before and after puberty to determine if any of the ob- served behavioral changes result from a change in catecholamine biosynthesis. Another exciting development is the appli- cation of the in situ hybridization histochemistry technique to study mRNA changes in individual cells. The technique is identical to immuno— histochemistry, except that it uses a DNA probe to identify a mRNA rather than an antibody to find an antigen. The major value of this tech- nique is that the researcher can see if there is a heterogeneous response in a set of cells (pro- ducing the mRNA of interest) to a specific stimulus. Such a situation is very likely to occur in a complex tissue like the brain. In essence, this technique will bring to the area of gene ex-— pression what immunohistochemistry brought to the area of protein/peptide expression in neural tissues. The area of behavioral genetics also appears ripe for exploration, using the new molecular approaches that are capable of studying genomic expression and regulation. Several strains of mice and rats and a large number of insect mu- tants already available are genetically deficient in specific gene products having direct in- volvement in behavior. The hpg (hypogonadism) mouse, which lacks luteinizing hormone releasing hormone (LHRH), and the tfm (testicular fem— inized) mouse, which lacks androgen receptors, are two examples of such vertebrate mutants. Such animals could be used as model systems to determine the involvement of the missing gene(s) products in eliciting certain behavioral states. Site of Action. In considering the behavioral actions of hormones, it is necessary to ascertain where hormones act. Special problems arise in the case of the peptide hormones. Since these hormones are often administered peripherally, it is essential to establish whether their effects are central or peripheral and whether peptides are capable of passing the blood-brain barrier. It is also important to determine whether the peptide of interest coexists in neurons with other bio- logically active molecules; in such cases, it may not be possible to mimic the action of a pep-— tidergic neuron by administering the peptide alone. Not all portions of the nervous system are equally sensitive to the behavior—inducing ef- fects of hormones. For example, the wandering 35 behavior in Manduca sexta can be evoked by bathing the isolated CNS with physiological concentrations of ecdysteroid. Although the hormone-sensitive sites that drive the behavior have been localized to the brain, they are dis- tinct from the centers responsible for the pat-— terning of the behavior itself. In order to be able to stipulate what hormones do to neurons, it is first essential to specify precisely which neurons are involved in the generation of a given be— havioral pattern. A system referred to earlier in which progress is being made is the egg-laying behavior of Aplysia. This animal possesses a paired cluster of neurosecretory cells, the bag cells, situated next to the abdominal ganglion. Excitation of bag cells or injection of bag-cell extract releases a 2— to 3-hour program of stereotyped behavior that results in egg-laying. (A variety of long-term electrical responses from identified neurons in the abdominal gang- lion are also evoked by these procedures.) The bag-cell factor that evokes egg-laying behavior has been isolated and sequenced. Immunocyto- chemistry utilizing antibodies prepared against ELH shows that all of the neurons in the bag-—cell clusters apparently contain the peptide. Cells showing ELH immunoreactivity are also sparsely scattered through other regions of the CNS. ELH does not account for all of the neural responses seen after a normal bag-cell discharge. Other peptides have now been isolated from the bag cells and have been shown to have selective actions on central neurons in the abdominal ganglion of Aplysia. ELH augments bursting in cell R-15; an inhibitory peptide causes prolonged inhibition of cells L3 and L6; and a third peptide provokes transient excitation in another population of identified neurons. All of these peptides are apparently released into the blood during bag-—cell activity. The role of each factor in coordinating the behavioral and physiological events that occur during egg-laying remains to be elucidated. The action of ELH has been studied with two neural preparations. The first is a set of iden- tified buccal neurons that may play a role in the inhibition of feeding behavior by ELH. The second is an endogenous bursting neuron, R-15, in the abdominal ganglion. The latter system is better characterized. ELH exposure enhances the amplitude of the pacemaker potential in R-15, an effect seen within 1 minute of addition of the peptide and persisting for a number of hours. The fact that R-15 is a very large neuron (cell body diameter greater than 250 microns) provides some opportunities for biochemical analysis of peptide action that are not possible with other systems. This potential has not yet been realized for ELH, but it has been for the biogenic amine serotonin (S—HT), which mimics some of the effects of ELH on R-15. The char- acteristics of R-15 have allowed a direct dem- onstration that S-HT effects are mediated ''through cyclic adenosine monophosphate (cAMP). More recently, the phosphorylation of specific R-15 proteins has been shown by preloading the cell with adenosine triphosphate (ATP) by in- tracellular injection and then stimulating the neuron with 5-HT. The relationship of these proteins to the conductance changes evoked by 5-HT is a point of great interest. New work on pituitary peptides is challenging the concept that these peptides enter the brain to exert direct neurotrophic action. Controversy exists as to the reliability of such phenomena as the enhancement of memory produced by AVP and the memory deficits observed in rats ge- netically devoid of AVP (Brattleboro rats). Also, the demonstration that a vasopressor antagonist analog of AVP can reverse AVP's behavioral ef- fects is challenging the concept of a direct CNS modulation of memory. Efforts are being made now to identify the site and mechanism of action for these peptide effects at behavioral and cel- lular levels. The hypothalamic peptides that control the release of pituitary hormones (hypothalamic re— leasing factors), having also been found in ex— trahypothalamic parts of the CNS, have been studied for possible CNS action. For example, LHRH has been shown to induce mating behavior in rats. The recent identification of cortico- tropin releasing factor (CRF) led to an attempt to identify a role for this peptide in addition to its classical hypothalamic releasing function. CRF has now been identified in extrahypothal-— amic brain sites; behavioral and physiological studies suggest a possible role in the CNS. The physiological and behavioral changes produced by CRF may reflect a CNS behavioral mobilization that parallels the physiological mobilization produced by the activation of the pituitary ad- renal system during stress. One might speculate that endorphins have a role in pain perception and the maintenance of well-being, or that AVP and CRF may regulate CNS mobilization to meet environmental chal- lenge and stress. Thus, peptides within the hy- pothalamic-pituitary-adrenal path may have complementary and dissociable functions, a hy- pothetical construct that can lead to specific testable hypotheses for behavioral function. Such a conceptualization raises the question of the degree to which peptides penetrate the blood— brain barrier. Little evidence exists to suggest that peptides such as the endorphins and AVP penetrate the blood-brain barrier in sufficient quantities to alter behavior. Indeed, these pep- tides disappear from the blood within minutes following subcutaneous injection. Determining the actual active compounds and the sites and behavioral mechanisms by which these periph- erally released peptides produce behavioral ef- fects are major goals for future studies. Autoradiographic mapping of cell receptors for behaviorally active hormones has made it 36 possible to generate a pool of candidate neurons that may have a role in hormonally activated behaviors. Where the neural circuitry for a be- havior is known, the pathway for producing the behavior can be compared with the location of hormone-concentrating neurons. Such compari- sons in vertebrates have revealed hormone-—con- centrating cells in brain nuclei at multiple loci in the behavioral pathway. Such loci include sen- sory structures, sensory brain nuclei, and the muscles themselves. Localized implants of hor- mones into specific brain nuclei can also acti- vate behaviors. Unfortunately, except in certain cases, it is not known whether hormone—con-— centrating neurons are the same neurons that respond to hormone application to produce be- havior. For example, in certain dedicated sys— tems (frog laryngeal motor neurons, bird syringeal motor neurons, rat bulbocavernosus motor neurons), the hormone—concentrating neuron is the behavioral effector. However, even in these cases it is not clear whether the steroid effect is due to a direct action on the neuron itself or on neurons (or muscles) to which the target cell is synaptically connected. Most identified pathways for reproductive behavior contain multiple hormone-target nuclei. If the behavioral actions of steroids arise via actions on steroid target cells, then a new set of questions arises. Are some hormone-—concentra— ting neurons more important than others in ac- tivating or in organizing behavior? Are some steroid—concentrating neurons holdovers from an earlier developmental period? There is now evidence that receptors for one hormone may be induced by exposure to a second hormone (for example, estradiol induction of the progesterone receptor). In some species, considerable dif- ferences are seen in the pattern of hormone- concentrating cells between males and females. Does neonatal hormone exposure determine the ability to concentrate steroids in adulthood? Alternatively, hormone-exposed cells might better be able to resist the ravages of histo— genetic cell death. Modulation of Neuronal Function. Steroid hormones studied in vertebrates have also been shown to influence neuron excitability, metab- olism, and morphology. The study of excitability began in the late 1960s and early 1970s. Estro- genic hormones have electrical effects in the basomedial hypothalamus and preoptic area that are consistent with their overall behavioral roles. Progesterone also affects the electrical activity of nerve cells, but whether it affects these nerve cells directly remains controversial. Androgenic hormones affect the nerve cells in the basal forebrain in a manner consistent with their behavioral roles, but whether those effects are due to testosterone itself or its metabolites has not been determined. In isolated frog spinal cord, androgens increase the excitability of mo- ''mw tor neurons controlling male copulatory behav- ior. Such an effect can be abolished by treat- ment with cyclohexamide, suggesting mediation by protein synthesis. Glucocorticoid hormones can alter the electrical activity of pyramidal cells in the hippocampus, where they are bound, but some of these effects may also depend on afferents to the hippocampus. Most effects of steroids develop slowly and this, together with the abolition of changes in excitability by protein synthesis inhibitors, has suggested genomic mediation. On the other hand, some very rapid effects of steroids on neuron firing have also been observed, suggesting that steroids may act at the level of the membrane. The mechanisms for effects on electrical ac- tivity are not known. Hormones could alter the number or function of ion channels or could otherwise influence important membrane pro— teins. Chemical techniques are now being used to study the effects of steroid hormones on enzyme activities, levels of cellular constituents, and neurotransmitter production and _ reception. Estrogen and androgens have been shown to in— fluence key enzymes in the synthesis and deg- radation of transmitters ranging from choline acetyltransferase to monoamine oxidase. The endocrine state.of the animal may also influence neurotransmitter receptor number or _ the availability of the correct isoforms of the hor- mones for binding to the cytoplasmic receptor. Which effects are primary remains uncertain, and the task of correlating specific changes with behavioral actions is barely begun. Another perspective in this area deals with how the presence of steroid receptors in a neu- ron affects behavioral activity. Briefly, the ki- netic properties of hormone receptors in the brain are notably similar to those in peripheral target tissues. A large number of comparisons using antiestrogens, antiandrogens, and physio-— logically weak hormone agonists correlate long— term receptor occupation with maximum behavioral activity. The possibility of direct membrane actions of steroid hormones is yet to be investigated systematically. Following re- ceptor occupation some electrical changes are clearly documented, but it is not yet obvious which of the chemical changes are in the service of these electrical changes. Blockers of gene transcription and translation can also block be- havioral expression when administered to ste- roid-sensitive neurons in a neural circuit for reproductive behavior. In studies of eclosion hormone using isolated CNS preparations, it has been possible to show that a 5-— to 10-minute exposure to the peptide causes permanent alterations in functioning. The observation that one simple reflex is perma- nently altered after peptide exposure suggests that, before eclosion hormone acts, the stimu- lation of a defined sensory neuron both excites an effector pathway and coactivates an inhibi- 37 tory neuron, which in turn blocks the response of the effector cell. Circulating eclosion hormone apparently shuts down the inhibitory neuron, so that the same stimulus now results in the normal effector response. Experiments using both in situ and isolated CNS preparations suggest that in- creases in both intracellular calcium ion (Catt) and cAMP are involved in the early action of the hormone. Another feature of interest is that the development of behavioral sensitivity to the peptide requires prior exposure of the animal to ecdysteroids. Thus, the eclosion hormone system has potential use as a model for how steroids and peptides interact in regulating the activity of specific neural circuits. The neurodepressing hormone (NDH) occurs in decapod crustaceans and has been purified and partially characterized. This peptide inhibits the spontaneous activity of a large number of iden- tified neurons. It is released in a circadian fashion and appears to modulate the daily ac-— tivity-rest rhythm of the animal. The use of isolated CNS preparations has revealed that the action of NDH can be reversibly blocked by ouabain, suggesting that the physiological changes induced in the neurons may be due to alterations in the electrogenic sodium pump. As with the action of eclosion hormone, derivatives of cyclic guanine monophosphate (cGMP) mimic NDH, whereas cAMP derivatives antagonize the action of the peptide. Development. From conception throughout life, the organism is exposed to varying concen- trations of hormones. It is therefore necessary to understand when the critical periods of hormonal action occur in relationship to specific devel-— opmental stages as well as in relationship to the development of specific behaviors. Invertebrate lifecycle changes are often accompanied by en— docrine-stimulated behaviors. Obvious examples are juvenile hormone, eclosion hormone, and ecdysone. In addition to its behavioral roles, ecdysone regulates aspects of the development of the nervous system. During metamorphosis, cell proliferation and neuronal outgrowth occur when ecdysteroids are present. In Manduca, a group of identified neurons acquire a dependency on the presence of ecdysteroids for their sur— vival. Their death follows the decline in the ecdysteroid titer at the end of metamorphosis. In vertebrates, hormones act on the neural effectors for behavior during major develop- mental periods. Very early in development, during so-called sensitive or critical periods, certain hormones act to alter the later behav- ioral capacities of the animal. This organiza- tional period is followed by the latent period of childhood. The emergence of hormone patterns associated with puberty or with the first breed- ing season initiates the activation of adult re- production. In most vertebrates, reproduction is characterized by fluctuations in hormonal levels ''associated with the onset of favorable climatic conditions for the rearing of offspring. Repro- ductive senescence is accompanied by decreases in fertility and by changes in the frequency of reproductive behaviors. The ability of neurons to respond to hormones in adulthood can be determined by the hormonal milieu of infancy. An example of this process can be seen in the sexual dimorphism found in the brains of a number of vertebrates (bird, frog, rat, hamster), which appears to involve both or- ganizational and activational steroid effects. A sexual dimorphism has recently been described in the human corpus callosum, but an endocrine dependence has not been demonstrated. In some insects, sexual dimorphism of sensory structures appears to be genetically determined. There is considerable sexual dimorphism in the CNS of vertebrates and invertebrates. Two ex-— amples are the sexual dimorphism described in the human corpus callosum and the apparently genetically determined sexual dimorphism of sensory structures in some insects. Differences in morphology range from the volume of entire brain nuclei, to the size of nucleoli, to the num— ber of synapses on dendritic spines. These dif- ferences appear to be established by the action of steroid hormones during development or in adulthood. Some morphological effects of hormones on neurons may be due to stimulation of dendritic growth. Thus, one sex may have more primary dendrites from one sort of neuron than the other. Organotypic culture of explants from neonatal mouse hypothalamus has revealed that estradiol promotes neurite outgrowth. Neurite outgrowth at an early stage could profoundly influence the activity of other brain regions through a domino effect. In order to demonstrate that steroids directly induce neurons to grow new dendrites, it will be necessary to isolate cells from synaptic influence. The assumption underlying most of this work is that an increase in dendritic area promotes the establishment and maintenance of new synapses, and that such synapses are critical to behavior or to neuroendocrine regulation. It has already been shown in some cases that steroids affect the growth of neurons in nuclei of a circuit controlling behavior. The hormonal basis of play and parental be— havior also merits renewed effort as does that of vocal behavior, all of which form a prominent part of the reproductive repertoire in many vertebrate species. Play is a component of the social repertoire in most primate species, in- cluding humans. This behavior occupies an in- teresting developmental time slot, occurring during the so-called latent period of childhood. In humans, the incidence of high—energy-— expenditure (rough and tumble) play is a sexually dimorphic characteristic that is sensitive to the endocrine milieu of development. Primates and rodents also show dimorphic play behavior; hor- 38 monal influences on play and reasons for the decrement in play behavior around the time of puberty are not fully understood. With respect to maternal behavior, more needs to be known about which hormones organize the capacity for expression of parental behavior. The clinical relevance of such studies is obvious. Indeed, one school of thought holds that courtship and copu- latory communications formed the evolutionary substrate for the development of human lan- guage. Some social vocalizations appear to be organized and activated by steroid hormones during development. Since there are a number of animal systems in which the neural pathways for the production of vocal behaviors are charac— terized, this area should be particularly fruitful. Homeostatic Processes In addition to generating higher mental func— tions, the brain integrates the emotional state with autonomic and endocrine functions to pro— vide the homeostasis for survival under condi-— tions of environmental stress. For example, the regulation of body water is important for every cell in the body, including the brain cells. Recent findings suggest that specialized brain areas in the hypothalamus can sense the concentrations of chemicals and water in the blood circulating through the brain (osmoreceptors), while other sensors in the heart and great blood vessels can detect blood pressure (pressor sensors) and blood volume (volume sensors) and transmit this in-— formation over pathways to the hypothalamus. This information is integrated in the hypothal-— amus and compared with set standards for ho- meostasis. Under appropriate conditions of environmental stress (dehydration, hypovolemia), signals are sent to the magnocellular AVP neu- rons for release of hormone from the posterior pituitary into the bloodstream and action on the kidney for water conservation. These same sig- nals can release a kidney factor (renin), which culminates in another circulating hormone (an— giotensin II) capable of stimulating further AVP release and drinking behavior. These same sig— nals can directly activate hypothalamic drinking neurons. A similar network determines when to stop drinking and releasing AVP based on a sen— sor—integrator—effector circuit. Disturbances in these sensor-integrator—effector loops may re— sult in disease states such as inappropriate drinking, excessive release of AVP, deficient AVP, and many others. New findings of molec- ular, cellular, intercellular, and behavioral studies increase understanding of homeostatic mechanisms and provide hope for the treatment of many of these disease states. The concept of the internal environment, first proposed in 1878 by Claude Bernard, brilliantly explained how animals maintain a number of vital life functions within very narrow limits in ''the face of wide and rapid fluctuations of the environment. Bernard used it to explain the precise regulation by mammalian species of their blood pH, blood sugar, and body temperature. In 1932, Walter B. Cannon coined the term homeo- stasis for these constancies and extended the concept to include regulation of food and water intake. Cannon was concerned mainly with reflex mechanisms in mammals. For example, with re— spect to thermoregulation, he focused on phys-— ical (shivering and vasomotor changes) and chemical (increased thyroid and adrenal gland output) means of heat production and loss. In psychology, homeostasis became the assumption underlying theories of drive reduction and was used to explain far more complex behaviors than those with which Cannon dealt. Within this framework, work on feeding, drinking, and thermoregulation proceeds on the premise that when an organism is in a need state, experi-— encing a physiological deficit, its behavior will be directed toward diminishing the deficit—in other words, such behavior is goal-directed. A goal is thought of as an ideal state the animal is trying to achieve. Translated into modern con- trol systems terminology, a goal is equivalent to a set point. A discrepancy between the set point and the actual state of some variable (blood pH, body temperature) generates a signal, which acts to adjust the actual status to the subjectively ideal status. In other words, the animal is seen as a biological servomechanism. Homeostasis in both simple and complex ani- mals can therefore best be studied in terms of a servocontrol system model. This type of analysis provides (i) opportunity for the integration of widely divergent approaches to a common set of problems, (ii) a nomenclature, and (iii) a means of generating specific testable hypotheses. These advantages accrue even if the model should prove to be incorrect. In homeostatic regulation by the CNS, the controlling elements consist of behavior, the autonomic nervous system (ANS), and the endo- crine system. These elements almost always act in parallel, but in some instances one or another element is predominant. Even if homeostatic regulation is not accomplished by a formal sys— tem of the servocontrol type, some of the scien-— tific tasks that can be fruitfully pursued include: (i) defining feedback signals, (ii) determining how feedback signals are processed by the nervous system, (iii) defining the nature of the control- ling elements, (iv) determining what might con- stitute a set point or similarly functioning device, and (v) determining where error signals might be generated and how such signals drive behavior and other controlling elements. It should be emphasized that not all motivated behavior is a consequence of direct homeostatic imbalance. Certain motivated behaviors reflect the operation of neural mechanisms that have evolved to anticipate possible homeostatic im- 39 balance. The study of these nonservocontrol mechanisms is also vital for a complete under- standing of homeostasis. Feeding Behavior. The investigation of in- gestive behavior is flourishing in three major areas—the ontogeny of feeding, the control of meal ingestion, and the relationship between food intake and body weight. New behavioral techniques permit the investigation of rat pups' feeding from birth until weaning. Neurobio- logical studies reveal that feeding and non-— nutritive sucking are under separate control. Further, 2-day-old pups can learn and remem- ber, are susceptible to positive reinforcement, and can make operant responses that lead to intracranial self-stimulation. The satiating ef- fect of food appears to develop about day 10, and there is evidence that the neurological mechanisms of satiation are active before the hormonal mechanisms are functional. Current analyses of the meal attempt to un- derstand the mechanisms that initiate, sustain, and terminate it. Among the most interesting recent work is analysis of the effects of gluco— privation. Receptors in the hindbrain are suf- ficient to mediate the response, and it has been shown that an episode of glucoprivation can ini- tiate feeding several hours later, thus suggesting the operation of a metabolic memory. Hypo- thalamic norepinephrine released during gluco- privation may mediate the feeding response in the region of the paraventricular nucleus. At the neurophysiological level, units have been found in the lateral hypothalamus and frontal cortex that respond to visual cues for food just prior to the initiation of feeding. In contrast to the experimental neglect of the positive reinforcing effect of food, its satiety effect has been studied extensively in the past decade. Significant findings include: (i) Ingested food activates satiety mechanisms from the mouth, stomach, and small intestine. (ii) Gas- trointestinal peptides such as cholecystokinin (CCK), bombesin, pancreatic glucagon, and so- matostatin have satiety effects. (iii) The satiety effect of all of these peptides except bombesin is abolished by abdominal vagotomy, which sug-— gests that they act peripherally. With the ex-— ception of the evidence in sheep for a satiety effect of brain CCK, there has been no sub-— stantial advance in knowledge about the central mechanisms of satiation. An interesting line of work concerns the pos-— sibility that the quantity and pattern of circu- lating amino acids control protein, carbohydrate, and total energy intake by affecting the uptake of tryptophan and tyrosine into the brain. The changes in uptake of these two amino acids are said to produce their ingestive effects by al- tering central serotonergic and catecholamin-— ergic mechanisms, respectively, which in turn alter food preference in subsequent meals. ''There is a large gap between work on control of meals and work on control of body weight. Currently, there are several suggestions for physiological and psychological mechanisms to bridge the gap. The physiological mechanisms include cerebrospinal fluid insulin, circulating glycerol, parameters of adipose tissue (such as cellular size and number, and lipoprotein lipase activity), and diet-induced thermogenesis. The psychological mechanisms include infant feeding experiences, cognitive control of eating (re- strained eating), and conditioned satiety and desatiation. The field has moved quickly to explore the possible involvement of the many new peptides found in the brain and the gastrointestinal tract and has obtained interesting results. Of the many peptides tested, CCK may be considered a par- adigm case; its satiety effect was demonstrated initially in the rat and more recently in both the lean and obese human. CCK's satiety effect in the human may be therapeutically useful for weight reduction. Other peptides with a satiety effect should move into clinical testing in the near future. A small cardioactive peptide (SCP) has been isolated and purified from Tritonia and Aplysia. In Tritonia, the main source of the peptide is the neuron B-11 in the buccal ganglion. Stimulation of this cell results in both peripheral effects (enhancement of the contractility of the esoph-— agus) and central effects (activation of the cen— tral pattern generator for feeding movements). The dual action of this peptide on the feeding system of mollusks is analogous to the dual ac— tions of CCK in mammals, i.e., involvement in regulating gastrointestinal tract function as well as in the central effects of satiety. Another feature of interest is that B-11 contains acetyl— choline (ACh) as well as SCP, although the functional relevance of this colocalization is unclear. CCK is the prototype satiety hormone: A peptide released from the gastrointestinal tract, it signals the brain to terminate feeding. Usually assayed by systemic administration of the pep-— tide with a subsequent decrease in feeding in hungry rats, recent evidence suggests that this effect is also peripherally mediated. However, some controversy exists as to the specificity of this behavioral effect (similar doses of CCK produce malaise), a problem that must be taken into account in the design of feeding experi-— ments. Evidence now shows that CCK is also localized in CNS neurons and, therefore, may also have a central site of action. Another gastrointestinal hormone, substance P, also has been localized to the CNS at many levels of the neuraxis. In the spinal cord, sub- stance P may have a role in pain modulation and in the effects of opiates on pain processing. In higher levels of the neuraxis, substance P may modulate dopaminergic function. 40 One of the strongly emerging themes is the importance of the abdominal vagus nerve for normal ingestive behavior and the satiety effect of gastrointestinal peptides. Most if not all of the effects of vagotomy are presumably the re- sult of loss of the afferent vagal fibers. This suggests an important role of visceral afferents for ingestive behavior. Unfortunately, very little is known about the neurophysiology and neuro- anatomy of these afferents. Such knowledge is needed to exploit the behavioral observations. In the CNS, the hindbrain is where taste af- ferents and vagal afferents interact, and it con— tains receptors for glucoprivic feeding. It appears to be a site of sufficient importance and complexity to merit intensive investigation, particularly through the use of the decerebrate preparation. The classic eating syndromes due to lesions of the lateral and ventromedial hypo— thalamus should be viewed as experimental challenges instead of neurological explanations. They can be analyzed only after the normal controls of feeding and body weight have been rigorously identified. The relationships between ingestive behaviors and other behaviors have been the subject of much recent ethological work. The controls of feeding and drinking, when they are available, will give researchers mechanistic leverage to determine which behavior becomes manifest when feeding or drinking is satiated. The rela— tionship of food satiety to sleep is the most ob- vious and interesting example. In this context, understanding is needed also of how ecological constraints can effect the size and frequency of meals. These same constraints may also have important implications for the assessment and treatment of disordered feeding in humans. Drinking Behavior. Work on drinking behavior has produced an unexpected development in the finding that centrally administered and pe-— ripherally administered angiotensin II probably have access to different receptor sites and serve different functions. Peripheral angiotensin II elicits drinking, but not salt appetite; central angiotensin II elicits both. The importance of central osmoreceptors for eliciting drinking in response to cell dehydration is supported by the finding in dogs that such drinking can be abol- ished by decreasing the osmolality of the blood perfusing the head. On the other hand, the re— ports that total abdominal vagotomy or selective gastric vagotomy markedly decreases or abol-— ishes drinking in response to the administration of hypertonic saline are difficult to comprehend. The interpretation of this result is complicated by the fact that abdominal vagotomy or some other mechanism, one not dependent on the ab- dominal vagus, is important for drinking after water deprivation. Although the mechanisms for drinking in response to hypertonic saline and hypovolemia ''Geo have been studied the most, the diurnal rhythm and the amount of food ingested are the main determinants of when drinking occurs and how much is consumed under conditions of ad lib water and normal temperature. The mechanisms for these effects have received much less at- tention. Therefore, the recent work on food— related drinking represents a new and promising development. Temperature Regulation. Progress in under-— standing temperature regulation has made it necessary to replace the old model of a neural thermostat in the anterior and preoptic hypo- thalamic areas with a new model, in which in- tegrators for thermoregulatory responses are located at many different levels of the neuraxis. According to the new model, individual thermo-— regulatory responses can be elicited or sup— pressed, while other responses are unaltered; there are many experiments that support this view. The impetus for this change came from experiments showing that behavioral heat seek- ing and avoidance remained intact, although thermoregulatory reflexes were absent or defi- cient following lesions in the preoptic area. It is obvious from these results that behavior must be considered in future studies on thermoregulation. The area of research on peptides and thermo-- regulation is extremely active. Arginine AVP reduces fever in sheep and lowers normal body temperature in rats. Bombesin, neurotensin, B-endorphin, and a host of other peptides also lower body temperature. Unfortunately, most of this work fails to emphasize the importance of behavior; only body temperature is measured, leaving important fundamental questions un- answered. When peptides alter body tempera— ture, it is essential to ascertain whether they act by shifting the set temperature, or whether the set temperature is unchanged and the peptides act by altering one or several thermoregulatory effector pathways. This question can be resolved by allowing an animal to work for heat or cold after peptide administration. If the behavior goes along with the change in temperature (if temperature goes up and the animal increases responding for heat, or decreases responding for cold), then it is possible to conclude that the peptide is acting on an integrative thermoregu— latory pathway. If the behavior compensates for the change in temperature, then the peptide is having a secondary effect. There are also non-— behavioral methods for determining set point changes, but an experiment that does not use some method for this determination reveals very little. Also worth noting is that administration of a peptide or any other endogenous compound can be viewed as administering a drug; therefore, experiments should be designed taking into ac-— count pharmacological and behavioral principles. Temperature regulation is intimately inter— connected with other behavioral systems. Ad-— 41 ministration of a drug or endogenous substance that alters feeding, drinking, or sleep is also likely to have an effect on body temperature. Work on the relationships between temperature and other regulatory systems is an important area for future research. The circadian rhythms of temperature need to be taken into account in work on homeostasis. Although mammalian body temperature is main— tained within narrow limits, it shows a circadian rhythm that has been rarely studied. Is the rhythm regulated? Can animals defend against thermal stress equally well at their peaks and troughs of body temperature? What components make up the entity we call body temperature? What happens when these are dissociated? What areas of the nervous system are involved in circadian temperature regulation? All of these questions need to be answered and are important topics for future research, particularly given the high correlation between body temperature rhythms and manic-depressive psychoses and moods in general. Reproductive Behaviors. The prominent role of the steroid hormones in vertebrate reproductive behaviors is well known. In the male, reductions in testicular hormone levels lead to reductions in masculine mating behavior. In the female, the highest levels of mating behavior occur at the time of ovulation. In ovariectomized animals, mating behavior can be induced by the serial administration of estrogen and progesterone. Variations in levels of gonadal steroids also cause behavioral changes other than those in— volved in mating, of which the most widely studied example is aggression. Across a wide range of vertebrates, large decreases in tes- ticular hormones result in significant decreases in aggressive behavior. The timing, nature, and object of aggressive responses appears to depend on environmental circumstances. There is also evidence that the hypothalamic peptide LHRH can activate female reproductive behaviors. LHRH promotes secretion of lutein-— izing hormone (LH) from the anterior pituitary, which in turn promotes steroidogenesis from the ovary. LHRH appears to affect reproductive behavior in a capacity distinct from its role as a releasing hormone, perhaps serving as a neuro-— transmitter or neuromodulator. Some evidence also exists for the presence of pituitary hor- mones in the CNS. A relatively unexplored possibility is that gonadotropic hormones them— selves may modulate reproducteive behaviors centrally, independent of their gonadal role. Highly seasonal species (birds, fish, lizards, am— phibia) are proving to be useful model systems for looking at gonadotropin effects. One somewhat neglected group of hormones in reproductive behavior is the prostaglandins. These substances, hydroxylated fatty acids de- rived from the metabolism of arachidonate, are ''found in many different tissues. Because of their rapid inactivation, it has been assumed that they can act only as local hormones. However, many studies (in rats, guinea pigs, hamsters, fish, lizards, and frogs) have shown that centrally and systemically administered prostaglandins of the E or F series can have powerful effects on re— productive behaviors. Another promising area for experimentation is pheromone research. Pheromones have been shown to be important in reproductive and other behaviors in invertebrates, and a growing body of literature indicates their significance in mam— mals, including primates. Sleep. Sleep in almost all mammalian species comprises two major states. One is rapid eye movement (REM) sleep, which is characterized by a low voltage electroencephalograph (EEG), the presence of rapid eye movements, inhibition of skeletal muscle tone, cardioréspiratory irregularity, increased cerebral blood flow, diminished hypothalamic regulation of body temperature, and several other major physio— logical characteristics. REM sleep is also char- acterized by distinctive patterns of neural activity, including increased bursts of neural discharge in the pons, lateral geniculate, and occipital cortex (so-called PGO spikes), de- creased firing rates of raphe neurons, and inhi-— bition of spinal motor neurons. Psychologically, REM sleep is associated with dreaming, although it is widely recognized that dreams and dream-— like mentation may also occur in the second major state of sleep, nonrapid eye movement (NREM) sleep. NREM sleep is characterized by the presence of high voltage, slow EEG activity and/or sleep spindles, and the absence of rapid eye movements. In general, physiological ac-— tivity (e.g., respiration, heart rate, blood pres— sure) is less variable during NREM sleep than during REM sleep. Sleep normally begins with the NREM phase. After approximately 70 to 90 minutes, the healthy young adult human enters the first REM period. (The period from the onset of sleep until the first REM period is termed REM latency.) Sleep then consists of a regular alteration be— tween NREM and REM phases. The amount of sleep, the time at which sleep occurs, and the relative ratios of REM and NREM sleep vary as a function of age. At birth, REM sleep occupies about half of total sleep and subsequently de- creases to the adult level of approximately 20-25 percent of total sleep. There is frequently a further small decline of REM sleep with ad— vanced age. In the infant, sleep gradually evolves from a polyphasic distribution to the well-known adult circadian rhythm of sleep-wake activity. Recent findings in the study of daytime sleepi- ness have indicated the presence of prominent ultradian sleep tendencies that become stronger during adolescence and constitute a major fea- 42 ture of sleep in the aged. There is remarkable phylogenetic variation in the total duration of sleep and in the length of the REM-NREM cycle. In general, small animals with high metabolic rates tend to sleep the most and to have the shortest REM-to-REM cycles. Several brain areas participate in the genera— tion of NREM sleep, including areas in the me- dulla, caudal pons, basal forebrain, and anterior hypothalamus. There is also strong consensus that pontine structures control the occurrence of REM sleep. Currently, there is a vigorous in- terest in determining the specific pontine areas involved, with special emphasis on the nucleus locus coeruleus, the ventral tegmental giganto-— cellular area, and the lateral tegmental field. There is little consensus about the major neu— rochemical regulation of NREM sleep. For many years, the serotonergic theory of NREM sleep enjoyed wide popularity. It is now generally re- cognized, however, that S-HT may have impor-— tant modulatory functions in sleep but is not the primary determinant of NREM sleep. Greater attention is now being given to a variety of putative neurochemical regulators, including possible endogenous sleep factors (delta sleep— inducing peptide, factor S, muramyl dipeptides, etc.), but no overwhelming consensus has yet emerged as to the necessity or potency of these putative regulators. On the other hand, there is strong consensus that REM sleep is initiated by a major cholinergic control mechanism. In addi- tion, much of the phasic activity of REM sleep (i.e., eye movement or PGO spikes) is inhibited by serotonergic mechanisms. Studies are needed on how ditterent neuro— chemically identified neuronal groups interact in the regulation of sleep-wake states. In recent years, there has been a strong emphasis on REM sleep and not enough on the regulation of sleep per se. Much is now known about neural dis— charge patterns during spontaneously occurring sleep states. The effects of putative sleep sub- stances should be compared with spontaneous patterns of the sleep cycle in order to determine their functional significance. Further research should be conducted into the basic mechanisms by which hypnotics, sedatives, and other drugs induce sleep or alter sleep architecture (i.e., lengthen REM latency or reduce REM time). The implications for these studies should be pursued in reference to basic pathophysiology of insom- nia, depression, etc. Animal models for sleep disorders should be pursued. For example, either chronobiological manipulations or pharmacolog- ical agents could be used to study the mechanism of short REM latency and narcolepsy. Because sleep is ubiquitous among mammals and reliably increases following sleep depriva— tion, it likely serves important biological func-— tions. Indeed, recent studies have shown that rats suffer severe pathology and death following prolonged sleep deprivation. However, the @a- ''es function of sleep is unknown. It is perhaps the greatest single gap in our understanding of sleep. There are several theories including those fea-— turing rest, tissue restoration, conservation of energy, restoration of brain synapses, protein synthesis, and behavioral adaptation. Theories of REM functions feature drive regulation, neural development, memory consolidation, program- ming, and unlearning. None of these theories of sleep or REM sleep function has gained a com- manding consensus in the field. Much greater emphasis is needed on sleep function. Otherwise, neurophysiological and neurochemical findings will exist in isolation from a comprehensive biology. Circadian Modulations. Human rhythms of body temperature, sleep, hypothalamic-pituitary hormones, and alertness and performance meas- ures have all received a great deal of attention. Under conditions of social and environmental entrainment in normal individuals, the timing of these processes produces a complex phase map of relationships. There are also important ul- tradian events of the daily neuroendocrine rhythms. The timing of such specific behaviors as sleep onset, lying down, waking up, exercise, and meal taking all have recently been shown to produce acute responses in physiological and endocrine functions, superimposed on both cir- cadian and ultradian events, and represent im- portant homeostatic mechanisms. Circadian modulations of physiological and behavioral states are produced by a circadian system consisting of meural and endocrine structures. It is likely that these structures form a hierarchically organized axis (analogous to the vertebrate hypothalamic—pituitary—gonadal axis), which conveys secondary properties to the cir- cadian system. There is good reason to think that primary circadian properties reside at the sub- cellular level in specialized pacemaking struc-— tures. The circadian system within the organism is normally synchronized (entrained) to the daily and seasonal environmental cycle by rhythmic environmental variables (primarily light-dark cycles). The phenomenology and environmental control of biological rhythms are well under-— stood; however, the regulatory and modulatory roles of rhythms in CNS function and behavior have not been well established. Circadian rhythms exhibit two essential fea— tures. First, they are endogenously generated and free-run with a period approximating 25 hours in the absence of environmental time cues. Second, under usual environmental conditions the light-dark cycle serves as a Zeitgeber, or en- training stimulus, to entrain the period of the endogenous oscillating system to 24 hours. Studies of control mechanisms of endogenous chronobiological processes demonstrate that certain genetic properties underlie the control of rhythmicity—that is, the ability to transmit both 43 period length and phase coupling to external cycles in a variety of organisms (single cells to complex cell systems up to and including the human being). Neuroanatomical studies demon- strate specific oscillatory control regions of the CNS—such as the suprachiasmatic nucleus (SCN), pineal gland, brain stem neurons (pontine reticular neurons, locus coeruleus, dorsal raphe, etc.}—and the effect of a wide variety of bio- logically active substances on period, phase, and amplitude of endogenous biorhythms. The formal properties of biological oscillators are now being addressed. These include concepts of mutual coupling among multioscillator sys- tems, pacemaker and slave relationships, wave- form analysis, internal synchrony, and states of desynchronization. Advances are being made in the development of mathematical analytical techniques based on oscillator theory and sto- chastic and time-series concepts, and in the de- monstration of application to invertebrates, tidal and lunar rhythms, seasonal and annual rhythms, photo-periodism, migration patterns, hiberna— tion, neuroendocrine functions, and the sleep— wake cycle of mammals, especially humans. Other generalizations can be made about the physiological organization of vertebrate cir- cadian systems. A large number of physiological processes expressing circadian rhythmicity are organized in physiological hierarchies synchro— nized by a central rhythm generator that is thought to coordinate the activity of other cir- cadian oscillators. In conditions free of external time cues, desynchronization of internal pace-— makers can occur; presumably individual pace- makers become uncoupled from each other and free-run with separate intrinsic periods, thus disrupting the normal phase relationship among internal circadian rhythms. Internal desyn- chronization of circadian rhythms has been ob- served in a variety of vertebrate species, including humans. A special case of desyn- chronization occurring in several species, and most studied in the Golden Hamster, is referred to as splitting. In splitting, the rhythm of loco— motor activity dissociates into two components, which free-run with different periods until they stabilize approximately 180 degrees out of phase with each other. This has suggested that a sys— tem of at least two coupled oscillators underlies the mammalian circadian pacemaker system. Splitting holds promise as an experimental sys- tem to investigate the general process of in- ternal desynchronization. In many _ respects, splitting resembles the internal desynchroniza-— tion that occurs in certain human illnesses. During the past few years, knowledge about circadian modulations has been applied to mental health problems. Some animal models (e.g., aging in mice and rats) are being studied, but much of the research is focused on nonlaboratory phase- shift problems, such as those seen in jet lag and shift work. Sleep deprivation combined with ''phase shifts are being examined as treatments in laboratories as well as in field studies. Although much of this work is descriptive, certain tenta- tive conclusions emerge. It appears that the aging process changes the amplitude of circadian rhythms, and that the amplitude of the rhythm may be a predictor of the rate of phase adjust- ment. Shift-by-delay occurs more rapidly than shift—by—advance, but there is considerable in-— terindividual variability in the rate of shift. There is a clear need for continuing study of components of the circadian system. This in- cludes: (i) identification of retinal elements in— volved in photoreception and transmission of visual information necessary for entrainment to the brain, (ii) elucidation of the location and organization of circadian oscillators, (iii) char- acterization of the neural interaction between components of an oscillator, and (iv) identifi- cation of mneurotransmitters and modulators produced by entraining pathways, components of the oscillators, and outputs from the oscillators. Such information would provide an important basis, both for the experimental manipulation of the circadian system and for therapeutic inter- vention in disorders of rhythmicity contributing to mental illness. Analysis of isolated circadian oscillators is another important direction for the field. One of the most significant developments has been the identification of specific structures that prob- ably contain endogenous, self-sustaining bio- logical clocks. These include the pineal gland of birds and lizards, the suprachiasmatic nuclei of mammals, and the eye of Aplysia. It is now pos— sible to isolate these structures, place them in a controlled environment deprived of entraining cues, and observe their capacity to generate 24-hour rhythms. Such in vitro systems can be used to answer questions about organization within the clock structure. Understanding the properties of isolated clock structures permits an examination of how they fit with other com- ponents of the circadian axis in the intact animal and leads to an appreciation of overall circadian organization. These clocks appear to have many charac- teristics of the intact organism, making them valuable models for studying the functions of 24-hour biological clocks and their molecular and biochemical basis. Using the techniques now being developed, it should be possible to identify specific proteins that exhibit rhythmic patterns of concentration. These proteins can be meas— ured using computer—assisted image analysis. When a protein of interest is identified, it can be sequenced, synthesized, and a cDNA probe can be made for the gene of interest. This opens the door to analyses of how the genes involved in controlling the production of these gene products are regulated and integrated into the circadian ~ oscillating system. Another approach would make use of genetic 44 methods by identifying and then studying strains of animals with genetically determined varia— tions in the circadian system. Such genetically distinct stocks could provide powerful tools to use in the study of the molecular basis of cir— cadian rhythms. The clinical extension of these studies would involve the search for the genetic basis of diseases related to or caused by appar-— ent alterations in the circadian system; the af- fective disorders are an example. A word about the importance of the compara- tive approach in chronobiology is in order. Although little is known about the cellular- biochemical mechanisms of circadian rhythm generation, what we do know supports the view that the mechanisms may be similar across a range of organisms. In both the eye of Aplysia and the avian pineal oscillators, cyclic nucleo— tides and the mechanisms involved in their reg— ulation appear to be closely connected to the rhythm-—generating processes. Even at the level of neurotransmitter involvement, there are similarities between birds and mollusks—-that is, S-HT exogenously applied has phase-shifting ef-— fects on the rhythm of locomotion in birds and on the rhythm of compound action potentials in the Aplysia eye. Hence, there should be in-— creased use of model systems that show promise of leading to an understanding of mechanisms, without the application of undue phylogenetic constraints. Finally, the future should see the development of dedicated chronophysiology laboratories with the potential for both basic studies of human behavior and clinical studies of mental health disorders. Future research on chronophysiology must build on major recent conclusions derived from studies of humans' internal timing systems. In particular, studies of human subjects are needed to analyze and compare the impact of specific entraining agents on the underlying pacemaker function. These could include such behavioral events as waking, meal timing, and acute physiological stimuli (exercise, showering), as well as specific psychopharmacological agents known to affect cyclic behavior, such as lithium. In addition, by structuring the timing of the human sleep-wake cycle, experimental testing of multioscillator mathematical models can be carried out in laboratory environments with complete control of external temporal and physical events. Application of the results of such studies to investigations of the major men-— tal disorders should be carried out in order to determine the extent to which chronobiological pathology applies to the 24-hour-—disordered symptom complexes. The specific disorders that should be studied include uni- and bipolar de- pression, anorexia nervosa, chronic alcoholism, progressive dementia, chronic insomnia, and drug-dependency sleep disorders. In fact, one appropriate use of drug therapy might involve the attempt to normalize the circadian system, ae ''in addition to using knowledge about the cir- cadian system to determine the appropriate time of drug treatment. Drug efficacy is modified by a multitude of biological factors that may manifest circadian rhythmicity in efficacy and toxicity. In the ideal case where the rhythms affecting toxicity and therapeutic effects clearly differ, one could op- timize drug efficacy and reduce toxicity merely by appropriate timing of administration. Such an approach could even be used for drugs with a relatively long half-life, where the peak levels, not the half-life, are responsible for the toxic effects. There are studies demonstrating that chronopharmacology can lead to refinements in treatments involving chemotherapy agents and synthetic cortiosteroids. Clearly, chronophar-— macology's greatest potential is with drugs hav— ing a narrow margin of toxicity. Although there is a good deal of experimental evidence to suggest that nonpsychoactive drugs should be given on a circadian basis (even though this is not being done), very few studies on the importance of circadian-based timing of drug administration for psychoactive drugs have been carried out. Some evidence does exist to suggest chronopharmacology may have a role in psycho- pharmacology. Circadian variations have been described in human and animal brain tissue and CSF levels of neurotransmitters, receptors, and enzyme levels—-all factors involved in the pro- posed mechanisms of drug action in humans. Furthermore, circadian rhythms in the efficacy of a variety of CNS-active drugs have been re-— ported, including barbiturates, amphetamines, chlordiazepoxide, ethanol, morphine, and halothane. However, many of these studies have been marred by the idiosyncratic methodologies and terminologies used. Other problems exist in many of the studies in their potential application to clinical pharmacology. Much of this work re— sults in reports of a chronopharmacological ef- fect without any examination of the specific circadian rhythms that might be involved. Ap- plication to psychopharmacology is especially challenging, since so little is known about the mechanisms of psychoactive drugs to begin with, and measuring therapeutic endpoints of these drugs is often quite complex. The chronophar-— macology of patients with abnormalities of cir-— cadian rhythms secondary to affective disorders provides a special challenge, because it is pre— sumed that the circadian rhythms affecting drug activity are also altered. Especially in such conditions, but also more generally, optimal pharmacotherapy would base drug timing not on external clock time but rather on the individual patient's circadian cycle. In relation to homeostatic mechanisms, work on neuromodulation appears to be especially promising. Neuromodulation encompasses a rapidly increasing collection of diverse phenom— 45 ena in which a chemical agent (e.g., hormone, CNS peptide, biogenic amine) influences elec-— trical activity or neurotransmission via other than conventional means at something other than a traditional synaptic junction. Hormones and other modulatory substances that bathe rela- tively large volumes of neural tissue may influence conventional neural activity and neu- rotransmission at many synapses and thus regu— late entire sets or circuits of neurons. The implications of such action for the gating or switching of behavioral states are only beginning to be appreciated. The time course of such events is also significant, because it is generally longer than the conventional millisecond time scale of synaptic neurotransmission and may even last for minutes, hours, or days through cellular mechanisms involving posttranslational modification of proteins (e.g., phosphorylation) or induction of gene products. Single-unit studies in freely moving animals are also a fruitful line of approach to under-— standing homeostatic mechanisms. In verte- brates, units in tissue culture and brain slice preparations may also lead to new insights into the afferent side of regulation. Brain transec— tions below the level of the hypothalamus (de- cerebrate preparations) are useful to studies of feeding, because taste afferents and vagal af- ferents interact there, and receptors for gluco-— privic feeding are found there. This approach is also useful to studies of temperature regulation, because animals without a hypothalamus still have many thermoregulatory responses available to them. Mathematical modeling and ecological ap-— proaches to regulatory behaviors are seen as potentially productive areas, as are studies of circadian rhythmicity. As previously noted, the study of circadian rhythms of body temperature is of potentially great importance. With the ap-— propriate use of implanted telemetry devices for long-term recording of body temperature in un- handled animals, the neural loci of the body temperature rhythms may be discovered. Aber-— rations in the rhythms produced through the use of lesions and neurotoxins may lead to an un- derstanding of the underlying mechanisms in cyclic affective disorders, where temperature disturbances are common. In general, circadian fluctuations in behavior should be considered in designs of homeostatic experiments. Recommendations for the Future Explicit examples of research needs and op-— portunities are replete throughout this chapter. Summarized below are those of the panels' rec— ommendations that deal specifically with hor- mones and neuropeptides in the control of behavior. @ Development of detailed descriptions of the ''neural vertebrate behaviors circuitry underlying Elucidation of steroid and peptide interac-— tions in the development and regulation of function of neural circuits Studies of gene expression changes mediated by hormones with an emphasis on the molecular mechanisms producing behavior Investigations of the cell biology of hor- mone-responsive neurons involved in be-— havioral events Elucidation of the mechanisms by which hormones act in relation to critical periods of development and the effect of hormone exposure on the biology of the developing neuron Studies of hormonal influences on behaviors during developmental stages (e.g., puberty) Development of new approaches to under-— stand the endocrine-mediated behavioral changes associated with aging Studies of the production and release of 46 more than one transmitter (cotransmission) in order to characterize colocalized neuro— transmitters and their mechanisms of functional interaction Development of mutant animal species and transgenic animals for use as model systems to determine the involvement of the spe— cific gene product(s) in eliciting certain behavioral states Explication of the role(s) of neurotrans- mitters/neuromodulators in homeostatic mechanisms Delineation of sleep function: the regula- tion of normal sleep; the interaction of different neurochemically identified neu- ronal groups in the regulation of sleep-wake states; the identification of endogenous sleep factors; and the physiological function of sleep as related to behavior and homeostasis Determination of how various drugs affect the circadian system and development of animal models to investigate the neuro- biological bases and neuropharmacology of disordered biological rhythms ''CHAPTER IV BEHAVIORAL ANALYSIS OF PSYCHOACTIVE DRUGS Modern psychopharmacology began in the 1950s with the discovery that drugs could be used to treat selectively schizophrenia, depres— sion, and anxiety. One of the first was reserpine, an antipsychotic and antihypertensive agent, which was found to exert its effects by reducing the concentration of monoamines in the central and peripheral nervous systems. These findings stimulated research into the neurochemical basis of drug action. Biochemical psychopharmacology in the 1960s studied the effects of drugs on pre- and perisynaptic processes such as neurotrans— mitter synthesis, storage, release, reuptake, and metabolism. This body of work revealed that many but not all psychotherapeutic drugs act through modifications of these processes. The study of the effects of drugs on tissues and cells has received great impetus in recent years with the advent of powerful new tech- niques. As discussed in preceding chapters, the brain uses chemical messengers to transmit in- formation from one nerve cell to others, through synaptic contacts between the cells. A variety of chemicals is involved, notably biogenic amines and peptides whose actions are mediated through specific receptors and an array of postreceptor mechanisms (see chapter II). It is currently be- lieved that many of the effects of psychothera- peutic drugs are mediated by the drug molecules occupying the receptors or altering postreceptor mechanisms, which are the same, or are related to, receptors involved physiologically with the normally occurring transmitters. A major task of behavioral pharmacology is to understand the basic processes that translate these molecular interactions into alterations in behavior. Almost all of the prototypical drugs used in treating mental disorders were developed by the application of knowledge generated in both pri- vate and public basic science research programs. Systematic behavioral studies in intact experi- mental animals was an essential component of these drug development programs. The devel- opment of the phenothiazines, the first useful class of drugs in the treatment of schizophrenia, is the classic example. Chlorpromazine, the first of the phenothi- azines, came from the private sector develop- ment of a research program aimed at finding a drug with primary central nervous system (CNS) effects. This work began with the testing of an antihistaminic compound known to exert CNS 47 side effects. The tests used to guide research in the development of chlorpromazine, based on the line of behavioral pharmacology research that began in university laboratories in the 1920s, were entirely dependent on systematic behav- ioral studies in experimental animals. Similarly, development of new psychotherapeutic drugs will require continued use of experimental animals. Behavioral Pharmacology Behavioral pharmacology is concerned with the effects of drugs on the transactions of a more or less intact individual--human or exper-— imental animal---with the environment. The ef- fects may be direct and immediate or delayed. Behavioral pharmacology is concerned with systematic studies of dose-effect relations of classes of agents whose neuropharmacological mechanisms of action are known, as well as agents that affect behavior by unknown means. The objective evaluation of behavioral changes should draw on a range of methods for assessing performance, unconditioned naturalistic behav— iors, and learned behavior involving both simple and the more complex associations character-— istic of higher functions in human and non human subjects. There is particular interest in the study of drug effects on signs and syndromes produced in experimental animals, particularly primates, by behavioral and pharmacological manipulations that are similar to human neuropsychiatric con- ditions. Factors influencing these effects—-e.g., age, previous drug exposure, genetic makeup, current and past environmental influences in- cluding nutritional status--are the concern of behavioral pharmacology. Two distinct experimental approaches are used in psychopharmacology. Type-—1 experiments use behavioral measures and procedures for analyz-— ing drug effects. Type-2 experiments use drugs as tools for understanding behavior. These two approaches are complementary and not entirely dichotomous. Thus, information obtained from type-2 experiments may lead to a better under- standing of drug-—behavior interactions derived from type-1 experiments. Moreover, type-2 ex- periments are most valuable when they employ drugs thoroughly understood in pharmacological terms. Advances made in neuroscience tech-— nology now also enable the use of direct CNS ''manipulations. These prerequisites for under-— taking type-2 experiments currently hold best for research on catecholamine systems. More research is needed using the type—2 approach, because an ultimate aim of behavioral pharma-— cology is to specify behavioral functions in neu— ropharmacological terms. Type-1 experiments use simple, often func-— tionally irrelevant, responses to measure such pharmacological actions of drugs as pharmaco-— kinetic and receptor characteristics. Commonly used in monoamine research, these can be termed type—la experiments. For example, the blockade of stereotyped movements is often em- ployed to measure the effects of neuroleptic drugs, and the induction of head twitches to measure those of serotonergic agonists. Although both useful and convenient, these indices tell little about the underlying functional substrates important for normal behavior. Somewhat re- moved from type-la experiments are another variant, type-lb experiments, that employ measures allowing some assessment of the effects of the drug on normal function. In this category are measures of spontaneous locomotor activity and consummatory and operant behavior and tests of sensory function. Their use provides essentially behavioral assays for drug action and toxicity and does not generally reveal the organization of the underlying behavioral processes. The type—1 approach thus leads to information that is essentially correlational in nature, the correlation being between the response chosen for study and either clinical potency or some neuropharmacological index such as receptor binding. Those studies correlating neuroleptic potency in blocking stereotopy with clinical ef- ficacy and in antagonizing dopamine (DA) re- ceptor binding would be a case in point. While providing useful information about neuroleptic drugs, this approach tells little about the func— tional effects of this class of drugs—notably, how they affect behavior and reduce schizo-— phrenic symptoms, and how schizophrenia may arise from an overactive DA system. An exactly parallel argument also can be made in the case of antidepressant and anxiolytic drugs (which are also believed to act at least in part via mono— aminergic mechanisms). The type-1 approach is used a great deal in drug-—screening, because it is well suited for the classification of compounds within an existing scheme. This approach may not lead to the de-— velopment of compounds that are clinically more specific or potent, because the arbitrary re- sponses used to quantify the drug effects may be far removed from the behavioral processes and neural substrates from which the clinical symp- toms emanate. For similar reasons, this approach will not likely uncover the nature and neural organization of behavioral processes in normal organisms. 48 The type-2 experiment, on the other hand, uses drugs to investigate behavioral processes. For example, a type—1b experiment might reveal evidence of a rate-dependent effect of a drug using operant procedures, an increase in punished responding, or an effect on retention of one-trial passive avoidance following posttrial adminis— tration, suggestive of a change in memory function. In these three cases, the type-2 ex- periments would ask questions about the nature of the behavioral processes underlying rate- dependent effects, and what the experiments on punished behavior can reveal about the nature of the behavioral mechanisms underlying such global constructs as anxiety and memory, re- spectively. Since certain drugs may have op- posite effects on consummatory and operant behavior, describing drug effects in terms of unitary motivational constructs ignores the com-— plexity of the underlying behavioral processes. Behavioral tests should be devised to isolate and characterize these processes more selectively, an approach that also eventually would provide a richer repertoire of type-—lb procedures. It is now possible to combine the technical develop-— ments in neuropharmacological manipulation and measurement of central neurotransmitters with development of more powerful and sensitive be- havioral paradigms to establish a specific role in behavioral control for a given neurochemical substrate. In this context, it is useful to review tradi— tional behavioral pharmacology as reflected in the study of the behavioral effects of amphet- amine-like drugs. Amphetamine is one of the most widely studied indirect catecholamine agonists, and understanding the behavioral proc— esses and neural substrates of its characteristic effects has remained an important goal of re-— search for both clinical and theoretical reasons. In common with most drugs when systemically administered, amphetamine has a multitude of effects that could arise from independent or common behavioral and neural mechanisms. Sufficient data on the drug have now been col- lected for hypotheses about its action to be framed. For example, rate-dependent analyses of amphetamine effects describe many of its ef- fects on operant schedules, and studies employ— ing the drug as a reinforcer, or observing its effects upon behavior reinforced by intracranial stimulation, have suggested that the drug spe- cifically enhances the impact of reinforcing events (negative as well as positive) on behavior. The most sophisticated form of this view main-— tains that amphetamine enhances the efficacy of conditioned reinforcers. Alternatively, taking into account many of the effects of amphetamine on both conditioned and unconditioned responding, some investigators have suggested that the drug increases the prob- ability of all responses exceeding some minimal tendency. Hypothetically, through behavioral ''competition this would lead to increased re- sponse rate within a reduced number of response categories, culminating in stereotypy. This hy- pothesis is an extension of the rate-dependency principle and operates at the level of response production. A final point of view would empha- size that the disparate nature of the effects of amphetamine e.g., in addition to those already mentioned, it has effects on aggression and eating behavior) arise from the drug acting si- multaneously at many different sites, including peripheral loci. Each of these positions has its strengths and weaknesses; only recently have the different accounts begun to be tested against one another. Besides continuing to assess the sensory and motor effects of drugs, the field of behavioral pharmacology can now arrange the contingencies of reinforcement so that an animal will behave in a predictable manner for extended periods of time, during which behavior can be challenged with drugs. For instance, methods are available for measuring drug effects on how well animals acquire new behavior. It is now possible to give a rat or monkey the task of making a sequence of responses—e.g., hitting panels in a predeter- mined sequence—in order to get some fruit juice. The sequence changes each day, the animal thus having to learn a new sequence each day. The animal's performance on this task becomes quite regular, making it easy to test whether a drug given before the session influences the speed with which the animal learns the new sequence. Particularly important to the development of behavioral pharmacology has been the study of behavioral mechanisms of drug action--i.e., at- tempts to discover which aspects of behavior are most relevant to the particular action that a drug has upon the organism. For example, in some cases, a drug appears to weaken the strength of the reinforcer being used to maintain the behavior. In others, a drug appears to make the reinforcer stronger. Conditioned reinforcers also can be changed in strength. A drug can weaken control by discriminative stimuli that have been modulating the animal's performance. It can reduce or increase the rate at which the animal responds. Contributing to the complexity of the enterprise is the fact that all of these effects depend on the original strength or rate of the variable being considered. Rate dependency, which underlies much of what on first analysis appears to be interpretable in other terms, is extremely important. Future directions in the area of behavioral methodology for use in psychopharmacological work should focus on applying the hard—won principles of animal behavioral pharmacology to human studies. Answers to the following theo- retical questions are central to an understanding of the behavioral effects of many drugs, espe-— cially the neuroleptics, whose neuropharmacol- ogical actions are generally opposed to those of 49 the psychomotor stimulants. @ Rate dependency. How pharmacologically specific are rate-dependent effects, given that they apply to several drug classes other than the psychomotor stimulants? What are the behavioral and the neural mechanisms underlying rate dependency? For example, does rate dependency arise as a secondary effect of impaired or altered stimulus con- trol, from a change in reinforcement ef- ficacy or behavioral competition, or from different combinations of these depending on the drug in question? © Reinforcement efficacy. To what extent are the rate-based measures of reinforcement efficacy secondary to rate-—dependent ef- fects? Are paradigms involving the choice of alternatives (e.g., the matching law) use- ful? To what extent do the putative changes in reinforcement efficacy of environmental events produced by psychomotor stimulants determine their own reinforcing properties? To what extent can the apparent effects of the drug on stimulus control or reinforce— ment efficacy be attributed to alterations in response output? e@ Stimulus control. What are the relative contributions of the stimulus properties of drugs and their direct effects on sensory input to instances of impaired stimulus control? How useful is signal detection theory (whereby the effects on discrimi- nation are described by two statistically independent parameters of discriminative sensitivity and response bias) for quantita- ting impaired stimulus control? @ Paradoxical effects of amphetamine. A prime example concerns the effects of stimulants on hyperactive children, where the relevance of rate dependency has hardly been considered. Furthermore, the lesson from animal work—that the apparently ben- eficial effects of amphetamine on per-— formance in certain situations are often counterbalanced by detrimental effects in other situations—has not been explored. On the other hand, the effects of amphetamine at more clinically relevant low doses in suitably appropriate behavioral situations have not been greatly studied in either normal or hyperactive animals. ® Quantification of the antipsychotic effects of neuroleptics in animals and humans. In clinical terms, perhaps the most important questions regarding neuroleptics concern the nature of the mechanisms, behavioral or pharmacological, that underlie the apparent need for chronic treatment before anti- ''psychotic effects emerge, and how and to what extent normalization of psychotic behavior occurs. Both questions might be tackled by appropriate animal experi- mentation. Ethopharmacology Among students of animal behavior, modern interest in brain—behavior relationships was spurred by the discoveries of the 1950s that fo- cal brain stimulation with electrical pulses could evoke integrated sequences of courtship and copulatory responses, threat and defensive dis— plays, attack and flight behavior, and sequences of foraging, predation, and food consumption. The initial anatomical studies focused on cir- cumscribed loci, or centers, that appeared to contain critical neural processes for specific classes of behaviors. This early concept was quickly replaced when new neural degeneration and histochemical methods began to unravel the extensive networks from which integrated be- havior patterns could be evoked. The history of psychopharmacological studies of social behavior—including sexual, maternal, affiliative, associative, aggressive, and defensive behavior patterns—-parallels in many respects the neuroanatomical studies. Initial efforts at- tempted to link specific classes of behavior to the activity of specific transmitter substance in discrete brain regions. The concept of neuro— chemical coding of behavior emerged from the early neuropharmacological and neurochemical discoveries on feeding, drinking, and sleep behaviors. Similar early efforts to identify aggressive monoamines and a sex neurotransmitter were considerably less successful; for example, pro- found skepticism developed toward the evidence for labeling brain serotonin (S—HT) a civilizing neurohumor for its role in sex and aggression. In the past dozen years, it has been recognized that even seemingly basic and simple behavior pat- terns—e.g., reproductive, agonistic, ingestive— are under complex neural control involving elaborate networks and can be influenced by many of the neuroactive amines and peptides that have been discovered so far. Although broader concepts of the multitransmitter control of behavior have been developed from the initial simplistic notion of neurochemical dualism, the narrower perspective prevails today in such formulations as of DA as activator and 5-HT as inhibitor of aggressive and/or sexual behavior. Even though many more neurohumors have been added to the neurochemical profile for aggres— sive, sexual, feeding, drinking, and sleep behav- iors, the basic research strategy has not been altered: Attempts are made to link specific be- havioral functions and malfunctions to the ac— tivation of specific neural receptors. 50 The foregoing characterization of past and present research strategies has been somewhat exaggerated in order to emphasize that the driving forces were advances in neurochemical and neurophysiological methodology. The un- proven premise has always been that the analysis of specific animal behavior patterns was satis— factory and valid. Also to be noted are two con— siderably more pragmatic psychopharmacological research traditions that have progressed, some— times in interaction with the more basic re— search on brain-behavior relationships. These are (i) animal behavior models for the purpose of drug development and (ii) research approaches to animal behavior modeling pathological human behavior. An illustration of psychopharmacological re— search on aggressive behavior from a drug de- velopment and screening perspective may be useful. Historically, an important goal was to determine whether drugs of separate therapeutic and chemical classes could selectively and spe- cifically alter aggressive behavior in contrast to nonaggressive behavior. This objective began to be pursued more than 20 years ago in an attempt to differentiate classes of CNS drugs. Experi- mental protocols were developed that allowed rapid testing of large numbers of drug candidates in laboratory animals. Prototypic and experi- mental drugs were given to pairs or trios of mice that were previously isolated, or to pairs of mice or rats that were subjected to electric shock through the grid floor of the cage, and aggres— sion rated on simple scales. A specific example from this era of research was the attempt to use a drug's capacity to suppress a rat's mouse— killing response as a screen for its potential as an antidepressant. This type of pragmatic in- vestigation, although still in use, reveals little about the behavioral, physiological, and neural processes that characterize aggressive behavior patterns. Psychiatric interest in animal aggression re- search has been kindled by the possibility of modeling pathological human aggressive behav- iors. Aggressive behavior is not only part of the symptomatology of several psychiatric and neu- rological disorders, but excessive, unusual out- bursts of aggression may represent a pathology all by itself for which no specific psychophar-— macological treatments are available. A prom-— ising effort in using a biologically more relevant model situation is represented by the work on the behavioral and neurological sequelae of ma- ternal or peer separation in monkeys. From the view of behavioral pharmacology, the literature on drug action and social behavior patterns has been disappointing, chiefly because the most-used approaches disregarded the unique characteristics of these behaviors, and because the traditional laboratory research protocols focused on situations and behavior patterns of dubious biological relevance and validity. Social ''behavior patterns are different from other be-— haviors studied in psychopharmacology in that (i) they consist of interactions between at least two individuals; (ii) they follow a defined temporal course, i.e., aggressive, sexual, maternal inter-— actions occur in episodic fashion; (iii) social be- havior is composed of a repertoire of behavioral elements that are typical for a given species; and (iv) these elements occur in predictable patterns. Characteristically, when the primary research objective is neurochemical and neuro- pharmacological, the measurement of social be— havior is relegated to a single, coarse index. The ethological approach has provided a framework and methodology that match the unique features of behavior patterns such as sex— ual and aggressive activities. With its origin in field research, modern quantitative ethology focuses on biologically relevant situations and behavior in exquisite detail. This detailed be- havioral analysis has prompted novel questions concerning the underlying neural mechanisms. The following recent examples on ingestive, sex— ual, and aggressive behavior illustrate the point. In addition to the current neurochemical and neuropharmacological research strategies for understanding the effects of drugs on food in- take, some researchers use a precise and sensi— tive analysis of feeding behavior as a tool to understand how drugs inhibit food consumption and adjust patterns of feeding. Based on the work of ethologists, procedures were employed to reveal changes in the fine structure of a rat's eating sequence. With these moment—to—moment records of an animal's behavior, it was possible to formulate behavioral characterizations of pharmacological manipulations acting via dif- ferent neurochemical systems. Behavioral methods to examine exploratory, sexual, and social behavior patterns in pharma-— cological and toxicological studies have been introduced. Based on ethological descriptions of behavioral elements in the wild Norway rat, a detailed quantitative record of a range of social and, particularly, sexual activities revealed profiles that were characteristic for specific hormonal, pharmacological, or toxicological manipulations. Studies of aggressive behavior have employed experimental situations in the laboratory that engender sequences of behavior closely similar to those seen in field conditions and have meas— ured and analyzed the behavioral elements in aggressive interactions in detail. The study of ageressive interactions presents the following formidable methodological challenges. The pat-— terns of aggressive behavior are species— specific; they are composed of a substantial number of postures, movements, displays, and communicative signals; they include interactions occurring in episodic fashion; epochs of intense aggressive behavior alternate with relatively quiescent periods; and these behavior patterns 51 are the product of at least two interacting combatants. A detailed temporal and sequence analysis of aggressive behavior detects unique drug effects with unusual sensitivity; at comparable drug dose levels, no significant changes can be seen by simply examining session totals. The rationale for the study of drug effects ina social context is simply stated: Most mammalian species, including humans, are social in nature. Animals that are removed from their social context may in fact represent unusual biological entities, yet behavioral, neurophysiological, and neurochemical effects of drugs are routinely studied in single animals. Among the many valid reasons for studying the effects of drugs, endogenous substances, and environmental toxins in socially intact animals, two appear particularly important. First, the effects of drugs on behavioral, neurochemical, and other physiological processes appear to be influenced by the past history of social inter-— actions as reflected by the role or status that an animal occupies in a social group. Second, many types and facets of social behavior are the pri- mary targets of drug action. For example, social status in a group of primates determines drug effects on several levels. It appears that certain status—linked behavioral, physiological, and neu— rochemical characteristics are substrates of drug action. Support for this conclusion is accruing: Dominant and subordinate tree shrews differ in terms of catecholamine synthesis; dominant macaque monkeys differ from lower ranking animals in gonadal and adrenal hormone secre- tion; and dominant and nondominant vervet monkeys differ in the way they metabolize 5-HT in blood and in cerebrospinal fluid (CSF) and in behavioral effects of S-HT precursors and reup-— take blockers. Most significantly, several recent studies highlight that drugs such as amphetamines, al- cohol, and benzodiazepines alter social behavior in a manner that was not predicted on the basis of existing evidence from single-animal re- search. For example, the low level of social in- teractions in subordinate monkeys failed to be altered over a wide dose range of alcohol or am- phetamine, whereas the relatively high level of social and aggressive behavior of dominant ani- mals was profoundly modulated in a dose-— dependent biphasic fashion. Dominant monkeys display a remarkable sensitivity to several be- havioral effects of dextroamphetamine, which is possibly indicative of status-—linked differences in the dynamics of catecholamine systems. The rationale for the study of drug effects on socially living animals is all the more persuasive insofar as various types of social behavior are specifically altered by drugs. However, in con- trast to the extensive pharmacological investi- gations into conditioned performance measures, motor stereotypes, and behavioral indices of ''memory and perception, little is known about drug effects on the wide repertoire of social interactions. In future research, it would be im-— portant for investigators with expertise in the analysis of social behavior to pursue experimen— tal work involving drugs, and for pharmacologists and neuroscientists to include detailed quanti- tative measurements of complex patterns of social behavior. One of the most significant issues in etho— pharmacological studies is that of experimental control. In contrast to the well-defined and tightly controlled variables of the experimental situations most often used in behavioral pharma— cology, research protocols for the study of drug action on social behavior resemble those for field research and typically do not stipulate ex- perimental control over the behavioral measures. The subtle nature of many types of social behav— ior often precludes experimental intervention. An appropriate research method must not de- stroy the social process it purports to study. The proximate causes of social interactions remain to be elucidated; the episodic nature of such be— havioral sequences is often labeled as sponta- neous. The question persists: What accounts for the moment-to-moment changes in social in- teractions in terms of situational events and biological processes? In spite of the potential pitfalls, observational methods represent the strength of ethopharma— cological studies. Instead of relegating obser-— vational methods to a preliminary step in dealing with a research question, the most profitable strategy will be to employ techniques that elim- inate observer bias; ensure accurate, objective, and reliable records; and enhance sensitivity, so that these methods are comparable to more ex— perimental and manipulative research. Such techniques are already available and continuously being improved. For example, por-— table microprocessors dedicated to data acqui-— sition eliminate many problems of recording, transfering, and analyzing observational data. Novel automated methodologies in data acqui- sition for social behavior, as well as satisfactory analytical techniques, are currently being developed. Electrophysiological Analysis Electrophysiological approaches may offer the level of analysis that is required for answering questions about the roles of neurochemically identifiable groups of neurons in physiology and behavior and in the mechanism of action of psychoactive drugs. Progress in understanding the neural bases of learning is testimony to the fact that this approach is applicable to what were previously considered intractable issues in the field of brain and behavior. Single cell re- cording has several desirable characteristics. 52 Indeed, single cell analyses are increasingly being used in studies of psychoactive drug ef- fects, attention, affect, and sleep—waking-— arousal states. Current technology allows single cell studies in unanesthetized and unrestrained animals, conferring even greater advantage in specifying the roles of these neurons in physiology and be- havior and in the action of psychoactive drugs. For example, these studies have dramatically altered the hypotheses regarding the roles of monoaminergic neurons in sleep—waking—arousal states and revised notions about the mechanisms underlying hallucinogenic drug action. It is in- structive to review the necessary steps and ex- amine the major pitfalls of such studies. At the initial stage, biologically active compounds that might act as neuroregulators are isolated, puri- fied, and identified. The precise anatomical localization of neurons containing these com- pounds is then established. The next step is perhaps the most critical. Once the cells have been anatomically localized to a particular brain region, the animals are anesthetized, and microelectrodes are placed into this area in an attempt to record single unit activity. Any success in this venture requires that the cells one intends to record from must be reasonably numerous, and it simplifies matters if they have relatively large cell bodies. It is to be hoped that, as in the case of the monoamine- containing neurons, the neurons in a particular cluster will all have similar discharge charac-— teristics, and these characteristics will be dis— tinctive enough to be described as a neuronal signature. It is also essential at this early stage that the neurons under study be unambiguously identified as containing the neurotransmitter in question. The most direct way of doing this is to mark the recorded cell with an intracellular dye and then to process the tissue with a procedure that highlights cells containing a particular neurotransmitter. Following this step, a variety of basic studies can be carried out in the anesthetized and/or immobilized animal. Cell activity can be exam- ined in response to, e.g., peripheral sensory in- puts, brain stimulation, or systemic drug admin— istration. Information regarding membrane characteristics and the nature and signs of syn- aptic inputs can be obtained from intracellular electrophysiological studies of these neurons. If further isolation of the neuron is necessary to study intrinsic activity or to eliminate multiple inputs that may influence drug responsiveness, activity can be recorded from tissue slices maintained in vitro. Further specificity of a pharmacological response can be obtained by microiontophoretic methods. However, strong caveats regarding the artificiality and non- physiological nature of these latter two prep-— arations should be heeded. A number of laboratories have now had suc-— ''cess in bridging the gap between recording the activity of neurochemically identifiable neurons in the acute studies described above and doing so in unanesthetized and unrestrained animals. At this stage of recording in freely moving animals, a number of ideas must be kept in mind to make experiments maximally relevant. e At least some of the testing of these cells should be carried out with biologically re- levant stimuli and under biologically rele- vant conditions. e@ Drug studies should explore clinically rele- vant doses and treatment regimens. Our understanding of the effects of tricyclic antidepressant drugs on the brain was dra- matically affected when these drugs were administered chronically, as they must be in order to be clinically efficacious. In addi- tion, because drug effects are being asses— sed in behaving animals in these paradigms, any untoward side effects are more likely to be discovered. Furthermore, since individual animals are typically studied for several months, and because the same cell can often be studied for several days, this preparation allows experiments involving long-lasting drug effects and/or long-term drug treat- ment regimens. ® Conditioning paradigms should be employed in order to bring the experimental variables under strict control. ® In drug studies, any change in unit activity should be measured against the appropriate baseline. This will help discriminate a drug effect that changes behavior, which in turn changes unit activity, from an effect of the drug directly on the unit under study. It is to be hoped that, in the future, electro— physiological studies of neurochemically identi- fiable neurons can be linked to studies reflecting the action of these cells on their target neurons. These experiments could take several forms, e.g., electrophysiological studies of neurons in these target sites; neurochemical studies em- ploying on-line measures of neurotransmitter release or metabolism at the target sites (e.g., voltametry, push-pull cannulae, or brain tissue dialysis); or some analysis of functional change in an output system known to be densely inner-— vated by monoaminergic neurons (e.g., the mas— seteric reflex, which is a monosynaptic brain stem reflex that can be examined in freely moving animals). Neurobehavioral Pharmacology Major advances have been made in under-— standing the molecular mechanisms operating at 53 the synapse, yet functional studies have not capitalized on this information to generate pre- dictors about the relationship of receptor acti- vation to behavioral output. Nor have students of behavior generally accepted that multiple neuro- transmitter interactions are the rule rather than exception. Theories of neurotransmitter function focused on unitary roles of chemical messengers are likely to be wrong. The research on the ef- fects of brain catecholamines and neuropeptides illustrates the progress that has been made in ascribing behavioral functions to selective neu-— rochemical lesions. Biogenic Amines. The early emergence of varied and sophisticated tools for analyzing functions of the central catecholamines, DA and norepinephrine (NE) made the area a test-bed for functional investigations of more recently discovered neurotransmitter systems. In par- ticular, the development of histochemical fluorescence and other advanced, newer tech-— niques has enabled a detailed specification of the neuroanatomical organization of the central catecholamine projections. Drugs also exist for intervention at several stages of the catechol- amine neurotransmission process including syn- thesis, release, reuptake, and degradation. There are available specific agonists and antagonists for different receptor subtypes, as well as rel- atively selective neurotoxins. Sensitive assays and methods for measuring turnover have been developed, and many of the more recent devel- opments in neuropharmacology—e.g., in vivo voltametry—-are pointed toward monoamine research. Despite this array of techniques, functional analyses even in the catecholamine field have made only limited progress. Use of the neurotoxin 6-hydroxydopamine (6-OHDA) has provided some of the most im- portant evidence regarding functions of the central catecholamines. It is well established that intracerebral injection of 6-OHDA can lead to substantial and selective catecholamine loss with few signs of other forms of neurotoxicity. Using intracerebral 6-OHDA, it has been shown that damage to ascending catecholamine path- ways to a large degree contributes to the lateral hypothalamic syndrome. The different functions affected suggest that the catecholamines have general, nonspecific functions in behavior. This view has been bolstered by demonstrations of transient behavioral recovery through nonspec- ific stimuli (such as tail—-pinch), as well as by catecholaminergic agonists. However, this ap- parent nonspecificity has emphasized, rather than detracted from, the importance of these neurotransmitters in behavioral contexts and their mediation of fundamental behavioral and physiological processes that underlie most forms of behavior. What is the nature of these processes, and which catecholaminergic projections are in-— ''volved? Only limited progress has been made on the first part of the question. Broadly speaking, the behavioral unresponsiveness of the lateral hypothalamic syndrome has been linked to striatal DA loss. More tentatively, its transient somnolence and electrocortical synchrony have been linked to dorsal noradrenergic bundle dam— age, and its impairment in response to specific physiological challenges has been linked to dam-— age to noncatecholaminergic cells intrinsic to the hypothalamus. More recent advances have suggested that damage to the different terminal DA fields has different functional effects. Thus, 6-OHDA lesions of nucleus accumbens produce no inges— tive and postural deficits. The lack of behavioral responsiveness (neglect) can be produced selec-— tively to events in one part of space by unilat- eral striatal lesions, and paw preference in the rat is also affected by such unilateral damage. In contrast, unilateral damage to DA terminals in nucleus accumbens does not produce obvious behavioral asymmetry. This distinction between effects of 6~OHDA lesions in these different DA terminal fields has been further substantiated by demonstrations of dissociation of different as- pects of the behavioral response to ampheta-— mine. The latter, essentially type—1 experiments, can lead to conclusions relevant to the type—2 approach. For example, if most behavioral ef- fects of amphetamines are shown to depend on central DA projections, this would indicate that these effects result from a small set of under- lying behavioral processes. There are indications that lesions to other DA terminal fields may also have dissociable ef- fects. For example, DA depletion from frontal cortex is reported to produce selective deficits in delayed alternation, thus mimicking some of the effects of ablation of this region. Such re- sults raise the question of whether DA has par- allel functions in different forebrain structures leading to different effects following depletion that depend upon the input/output character- istics of these innervated regions. In general, there is a need for more sophis- ticated behavioral paradigms to characterize the neglect phenomena. These phenomena seem un— likely to result from primary sensory or motor dysfunction. Rather, the distinction would ap- pear to be between attentional failure (where the appropriate stimuli are not made available for triggering response selection) and activa- tional failure (where response selection may have occurred but cannot be executed). Methods are available for dissecting these possibilities; some have been employed in animal CNS re-— search but not yet directed toward the question at hand. Improved behavioral paradigms are also needed to compare functional deficits produced by depletion of different forebrain regions. This argues for an application to psychopharmacology of suitably modified neuropsychological tests of 54 different forms of behavioral processes such as memory and attention (see chapter V). Noradrenergic systems provide an intriguing parallel with the DA systems, inasmuch as 6-OHDA lesions of dorsal and ventral bundles produce different, sometimes opposite effects. Damage to the ventral bundle, leading mainly to depletion of hypothalamic NE, leads to weight gain, increases in feeding, proceptive sexual be— havior in females, and general reactivity; dorsal bundle lesions have none of these effects. In contrast, the dorsal bundle from locus coeruleus to hippocampus and neocortical areas has been implicated in several CNS processes including reinforcement, arousal, anxiety, and selective attention. However, it has not proven possible to provide conclusive evidence in favor of any one of these hypotheses or to specify their interre— lationships in any detail. This is because those effects of dorsal bundle lesions that are reliable do not obviously reflect a single process. For example, such lesions have been reported to produce effects as diverse as reductions in neu- ronal plasticity in visual cortex and attenuation of blocking of classical conditioning. Once again, these effects may reflect the functions of the diverse terminal areas rather than heterogeneity of dorsal bundle function. The NE neurons in the locus coeruleus may become activated under certain conditions that result in simultaneous changes in neuronal firing in many different areas having varied functions. What is clear about NE lesions is that they do not produce large changes in response output as compared with DA lesions. Although NE systems are also implicated in general, nonspecific processes, these are obviously of a different type from those mediated by the DA systems and perhaps reflect processes more related to input than re- sponse production. Different behavioral para-— digms can be used to dissect varieties of such attentional dysfunction; nevertheless, it is not yet apparent how dorsal bundle deficits can be described in these terms. Technological advances will continue to spur interest in catecholamine studies. Improved methods for investigating neurotransmitter turnover have already established several im- portant functional relationships. Insofar as re— cent results indicate dissociation in the functions of the different DA projections, which may be relevant to the earlier dissociations revealed by selective lesions, detailed behavioral analysis is warranted. New techniques for investigating in vivo neurotransmitter activity, such as volta— metry, might eventually aid this enterprise. In several areas, electrophysiological and neuropharmacological findings have combined effectively and made imperative a functional analysis of the system in question. Perhaps the best current example of this is the regulation of locus coeruleus NE activity by presynaptic alpha—2 and enkephalin receptors. By using ap- ''propriate microinjection techniques, it may be possible to reversibly activate or depress dorsal NE bundle function, thus providing a much- needed alternative to lesioning techniques. The use of central microinjections might also facil- itate the analysis of the behavioral effects of drugs such as amphetamines, benzodiazepines, and neuroleptics by producing more selective and dissociable behavioral effects than may be pos— sible with systematic injection. During the past 20 years, the amine theory of depression has stimulated considerable interest in the role of brain monoaminergic neural sys- tems in affective disorders and in the mecha- nisms of action of drugs and procedures used to treat such disorders. Hypotheses related to the biological basis of depression have been derived from observations on the pharmacological ac-— tions of a variety of psychotropic drugs, in- cluding monoamine oxidase inhibitors, tricyclic antidepressants, and drugs that alter neuronal stores of neurotransmitter substances such as reserpine. The observation that certain tricyclic antidepressant drugs block the uptake and reten— tion of NE by noradrenergic neurons led to the hypothesis that depression was the result of in-— adequate release of NE at synapses in specific areas of the brain. However, no temporal re— lationship exists between the onset of clinical actions (which take days or weeks to develop) and the effects these drugs have upon the nor-— adrenergic uptake system (which appear after a single dose). Such studies have led to the evalua-— tion of changes in neuronal activity and possible functional changes in neuronal autoregulatory processes after chronic administration of anti- depressant drugs. Current preclinical and clinical research em- phasizes changes in receptor density or affinity, or both, to clarify the pathophysiology of the affective disorders and schizophrenia and the mechanisms of action of antidepressant drugs, lithium, electroconvulsive therapy (ECT), and antipsychotic drugs. For example, the possibility of increased S-HT2, alphaz-adrenergic, beta— adrenergic, or dopaminergic receptors. in affective disorders and schizophrenia is being studied with biochemical, neuroendocrine, and neurophysiological methods. Animal studies have demonstrated decreases in the number of beta— adrenergic, alphaj-adrenergic, or S-HT2 re- ceptors in specific brain areas following sub-— chronic treatment with most antidepressant drugs or ECT. An exception is the increased S-HT2 receptor found after ECT. Human stud- ies have also directly demonstrated receptor changes (e.g., decreased platelet alphaz re— ceptors following chronic tricyclic treatment) or obtained data consistent with receptor altera— tions (e.g., changes in the 5-hydroxytryptophan-— induced increase in serum cortisol) following chronic tricyclic or lithium treatment. Cholin-— ergic and GABA-ergic receptor mechanisms 5S have also been of interest in the understanding of the pathophysiology of affective disorders and schizophrenia. Studies integrating preclinical and clinical research to define biological changes present in untreated patients and the relation— ships of these changes to clinical state, clinical response, or side effects are among the best means available to study these major mental disorders and the drugs used to treat them or prevent their recurrence. The meaning of changes in receptor number as revealed by ligand-binding studies and Scat- chard-type analyses is equivocal without specific studies of functional consequences of the changes in receptor number and/or affinity. For example, changes in the activity of epineph— rine(E)-stimulated adenylate cyclase may occur without change in the number of beta-adrenergic binding sites. Such studies of the significance of alternations in receptor number and affinity should include an examination of the biochemi- cal, physiological, and behavioral consequences; this work should be carried out in humans as well as in various species of laboratory animals. Basic research on the effects of chronic drug administration on receptor number, affinity, and function has had significant impact on the hy- potheses currently being tested by clinical investigators. Popular strategies currently em- ployed include biochemical measures such as measuring the Bmax (maximum ligand binding) and Kp (ligand affinity) of alpha-or beta- adrenergic receptors on platelets or lympho- cytes, respectively; assessing physiological responses after administration of agonists (e.g., blood pressure response after clonidine); bio- chemical responses (e.g., the plasma 3—methoxy, 4-hydroxyphenylglycol (MHPG) response after clonidine); and neuroendocrine responses (e.g., the S-hydroxytryphophan-induced increase in cortisol and growth hormone). The data such studies provide will be helpful in clarifying which receptors contribute the most to the etiology of the affective disorders or the mechanisms of ameliorative treatments (which are not neces— sarily the same). Future work should be guided by questions relating clinically relevant phenomena to the numerous effects of chronic antidepressant and antipsychotic drug administration on receptor density and function in laboratory animals and humans. This work could utilize positron emis- sion tomography to assess receptor density in specific regions of brain before and during treatment and in periods of remission. Clinical studies of receptor responsivity can suggest which receptor changes correlate best with im-— portant dimensions of depression or mania (e.g., severity, diurnal symptom changes, suicide); which may be trait or vulnerability factors; which predict responses to treatment, etc. This evidence will help to identify those receptor changes noted in preclinical studies that are ''most clinically relevant. In addition, specific studies should be directed toward the develop-— ment and validation of means of assessing al-— tered receptor functions in patients, clonidine- and yohimbine—induced changes in plasma MHPG and growth hormone, platelet transport of 5—HT, and platelet tritiated imipramine binding. Also, greater emphasis should be placed on studies that attempt to determine mechanisms of drug action. Studies concentrated on measurements of receptor number and/or affinity and indirect measurements of receptor function alone are less than optimal, given the current range of measurable parameters. More specific agonists and antagonists for dopaminergic, serotonergic, noradrenergic, and cholinergic receptors will be of value for studies of the role of specific neurotransmitters in reg- ulating brain and bodily functions, for binding studies, for treatment, and for challenge studies in humans. Such development will, e.g., facili- tate work on the role of S-HT in affective dis— orders and schizophrenia. Research on the interaction of neurotransmitter systems is en— couraged, as is study of the reciprocal effects of the endocrine systems on aminergic or pepti- dergic neuronal activity. Finally, note should be made of interactions between catecholamines and comparisons with other neurotransmitter systems. The behavioral functions of the catecholamines need to be dis— tinguished adequately from those of other neuro- transmitter systems. For example, ascending serotonergic projections, as well as the dorsal NE bundle, have been implicated in anxiety and in the anxiolytic effects of benzodiazepines. Any comparisons should include several tests of anx- iety, since this construct may reflect several processes. Damage to the cortical cholinergic projections from the basal nucleus of Meynert is implicated in the behavioral disabilities of Alz- heimer's disease. This contribution must be differentiated in behavioral terms from that of damage to the coeruleal—cortical innervation, which is also implicated in this disease. Neuropeptides. One of the first pieces of evidence suggesting a behavioral role for pep— tides came from studies of pituitary—adrenal function, demonstrating that the behavioral effects produced by removal of the pituitary could be reversed by adrenocorticotrophic hor- mone (ACTH) and other peptides such as vaso-— pressin (AVP). Later work established that these peptides could produce profound effects on ex- tinction of learned responses in normal animals, even with synthetic analogs reputedly devoid of endocrine activity. Variously interpreted as actions on attention (ACTH) or on memory con- solidation (AVP), these effects were hypothe- sized to be mediated via some neural substrate. Where this substrate is located and how these pituitary sources of peptide normally interact 56 with it remain largely unknown. A separate line of research unrelated to the pituitary studies led to the discovery of a pos- sible peptide signal in the initiation of drinking behavior and thus provided evidence for CNS action of a peripherally released peptide hor-— mone. Eventually, it was shown that angiotensin Il caused drinking when infused intravenously in microgram amounts, and that much lower doses (nanograms) than those applied intravenously also produced drinking when injected intracra— nially, suggesting that systemically generated angiotensin acted directly on the brain to induce thirst. More recent work has pointed to a ven— tricular route of access to sensitive sites in the region of the subfornical organ, and this infor- mation subsequently has been thought to be sent to the median preoptic nucleus for the mobili- zation of thirst. This elegant story of a blood-borne peptide involved in homeostatic function has been an impetus for studies of parallel mechanisms for feeding. Cholescystokinin (CCK) causes a dose- dependent decrease in food consumption in sev- eral animal species and has subsequently been suggested as a satiety hormone. Perhaps the single most important discovery of peptides as putative neurotransmitters 1lo- calized within the CNS has been that of opioid peptides as possible endogenous ligands for the opiate receptors identified in the brain. Opioid peptides have subsequently been implicated in all the classical effects of opiates such as their analgesic and hedonic action, as well as in nor- mal functions of pain, reward, learning, and emotionality. Substantial efforts also have been directed at exploring a possible role for these peptides in psychopathology and drug abuse. Recent work has begun focusing on the sites and mechanisms of action for exogenous opioid peptides, using mainly such techniques as local central injection of the peptide and opiate an— tagonists such as naloxone to block opiate recep— tors. The actions of exogenous peptides or their antagonists are taken to infer the sites and func— tions of their endogenous counterparts. Current areas of intense effort center on the identifi- cation of roles for both central and pituitary endorphins in stress-induced analgesia, memory, psychopathology, and opiate addiction and the identification of the function of whole families of peptides with similar but not identical receptors. Studies exploring the behavioral roles of pep-— tides have been increasing at a rapid pace in recent years. Some of these peptides were orig- inally characterized by their hypothalamic-— pituitary-adrenal axis function and subsequently have been hypothesized to have related CNS action; others, such as substance P, CCK, and vasoactive intestinal peptides, are gastroin— testinal hormones that now have also been lo- calized to CNS neurons (see chapter III). ''A wealth of behavioral pharmacological data and conceptualizations exists with which the effects of new peptides and their analogs could be compared. Initial characterization could direct future studies to systematic behavioral evaluation using established operant and etho- logical tests; this information is essential for characterizing and explaining their behavioral effects. For example, some peptides originally characterized by their specific endocrinological actions may be important candidates for CNS function in biologically relevant species—specific behavior. Development of specific peptide neurotoxins and antagonists is needed. Currently, antagonists are available for leutenizing hormone releasing hormone (LHRH), AVP, opioid peptides, CCK, and substance P. Future studies should be di- rected at characterizing the effects of these antagonists on specific aspects of behavior where other evidence has suggested a role for a particular peptide. The development of specific neurotoxins would provide an even more power-— ful tool for dissecting the functional relationship within different systems. Most brain peptides have endocrinological ef- fects in the periphery or act in some way within the hypothalamic pituitary axis. Critical studies are needed to identify the behavioral role (if any) of such peripheral actions of peptides. The relevance of these potential visceral signals for emotionality, motivated behavior, discrimination of state, and even learning and memory is an intriguing but largely untapped area. From the standpoint of clinical relevance, the resolution of sites and mechanism of action of peripherally derived or peripherally administered peptides is of great importance. Recent work has established that a variety of different peptides can influence activity in es— tablished neurochemical circuits (e.g., mesolim— bic DA system and opioid peptides, neurotensin, substance P). Future work will be needed to characterize the behavioral significance of these interactions by way of describing both the cel- lular and system-level function of peptides. In addition, an exciting potential area of peptide- transmitter interaction is suggested by the evi- dence of colocalization of peptides in neurons containing other transmitters (e.g., localization of CCK in mesolimbic DA terminals). At some future point, neuroscientists describ-— ing the behavioral effects of specific augmenta-— tion or destruction of a given peptide circuit will need to integrate these findings with cellular information about the nature of activity in the circuit. Such activity would include, for exam-— ple, the antecedent conditions for activity in the circuit, consequences of activity in other cir-— cuits, self-regulation at the cellular level, and more. This integration of molecular, cellular, and behavioral levels of analysis is necessary to characterize the role of peptides in brain 57 function. In peptide work, a wide gap exists between the identification of behavioral effects and the identification of sites and mechanism of action. For example, few firm hypotheses exist to ex- plain the plethora of subtle effects of systemic administration of peptides in learning situations. Indeed, some argue that these effects are medi- ated by peripheral receptors. Studies should be undertaken to determine whether systemically administered peptides do cross the blood-brain barrier, whether nonspecific physiological ef- fects do contribute to these behavioral changes, and where—inside or outside of the brain—such peptide actions are elicited. Finally, little effort to date has been directed at examining classical behavioral pharmacolog- ical principles in evaluating peptide actions. Studies are needed to explore factors known to determine drug action, such as the schedule of reinforcement, the nature of the reinforcer, and the discriminative stimulus properties of these compounds. Also largely unexplored are studies of the interaction of behavior and the history of the organism on the release and utilization of peptides. What environmental events or inter- action of past history and current environmental events activate peptide systems, and what role do these peptides have in the behavior that results? Drug-Induced Models of Psychopathology Animal models serve as bridges between studies in humans and in vitro studies. The ani- mal model can serve as a dynamic medium in which to test hypotheses about human psycho- pathology derived from in vitro studies or from human clinical models of disease states (see chapter V). It can also serve as a useful tool for predicting drug and other treatment responses. The following is a brief discussion of drug- induced models of human psychopathology found useful in psychopharmacological work. Modeling is, in general, a process of looking for analogical invariants. The mature model or an ideal animal model should satisfy the fol- lowing six conditions: 1. Similarity of behavioral state exists be- tween animal model and human model of the psychopathology or the human psycho- pathology itself. 2. The model has neurobiological mechanisms common to those of the human psycho-— pathology or can further elucidate these mechanisms (chemical, neurophysiological measures). 3. The behavioral state of the model can be altered by clinically effective treatment techniques. ''4. The model has predictive value for the development of new treatment techniques and new drugs. 5. The model has been or can eventually be produced in animals closely related to man, such as primates. 6. The model has inducing (both experimental and natural) conditions, common to the human condition. Perhaps the model most widely exploited both experimentally and theoretically in the current era involves the stimulant drug-induced psy-— choses in man. A brief history of its development will illustrate the step-by-step maturation of this theme in modern psychopharmacology. The amphetamine psychosis was differentiated from other toxic—hallucinatory states and documented as a hallucinatory—delusional syndrome in a state of clear consciousness in the early 1950s. The occurrence of primary delusions and Schneider— ian first-rank symptoms of schizophrenia are features of amphetamine psychosis that even today confuse the attempts of experienced di- agnosticians to differentiate it from paranoid schizophrenia. For example, in the early 1970s, as many as 15 percent of unsuspected new patients in State hospitals were later found to have phenylethylamine metabolites in their urine. The syndrome develops in an orderly sequence of emergent symptoms. An instructive feature to follow is the induction of delusions of parasit— osis. The gradual entrainment of picking and examining stereotypes preceded the rather sud— den development of the delusions of parisitosis. Administration of stimulants to former abusers under laboratory conditions has confirmed the drug-induced origin of the phenomenon. In a parallel series of developments, researchers first demonstrated that acute administration of stim— ulants and direct DA agonists to animals induced behavioral stereotypy. Pharmaceutical industry researchers then utilized the blockade of apo- morphine (a DA agonist)—-induced nausea in the dog as well as stereotypy in rodents to screen for the current series of popular neuroleptics. The EDso (dose effective in 50 percent) for blocking amphetamine or apomorphine stereotypy is highly correlated with the clinical potency for neuroleptics. Metabolic studies of DA and its metabolites in schizophrenic (but not pure paranoid) populations have not sustained the analogy, although an es— cape clause is afforded by the division of the DA system into its striatal and limbic components. The exceptions to the blocking of apomorphine or amphetamine stereotypes (such as clozapine and sulpiride) suggest that some newer, atypical neuroleptics may work selectively on the limbic rather than the striatal component of the DA 58 system. As early as 1973, it was pointed out that the chronic stimulant model evolved to a stage where tolerance developed to stereotypy, and late-stage hyperreactive behavior emerged that was analogous to the development of the human condition. The emergent behaviors included hy- perstartle limb flicks, head jerks, abortive grooming, and hallucinatory-like responses. These were the behaviors that were more dra- matically inhibited by drugs like clozapine than specific antidopamine drugs like pimozide. Many of the late-stage emergent behaviors inhibited by the new antipsychotics have been found fol- lowing acute lysergic acid diethyamide (LSD) administration associated with a reduction in brain 5-HT and its metabolite by 30 to 40 per- cent. Other neurobiological findings associated with emergent hyperreactive behavior syndrome include a decrease in tyrosine hydroxylase and DA cell body auto receptor supersensitivity. Recent attempts of animal modeling to simu-— late human clinical phenomenology have focused on symptoms rather than syndromes (see chapter V). Among the models currently used for study-— ing antianxiety drugs, behavior suppressed by a variety of means, such as electric shock as first shown in punished behavior procedure, has been the most useful. Many of the original measures were effected by anxiolytic drugs in parallel with their clinical anxiolytic efficacy such as sedation, muscle relaxation, anticonvulsant ac- tivity, or loss of motor coordination. These early tests screened most early anxiolytics including meprobamate. The later developed monkey—tam- ing paradigm and punished behavior model were the first behavioral measures that more spe- cifically correlated with the antianxiety effects of the drugs. Sedation, muscle relaxation, and ataxia properties of anxiolytic drugs are con- sidered exclusion criteria of side effects; thus we see the evolution of a mature animal model. A different approach, starting from a series of observations on lower primates, has led to a model of anxiety-panic disorders based on in-— creased locus coeruleus neuronal firing. Monkeys stimulated in the locus coeruleus develop intense behavior suggestive of apprehensiveness. This model has been helpful in developing new treat- ments (clonidine) of the anxiety associated with narcotic withdrawal. Recently developed is a human, as well as lower primate, model of anx- iety panic using acute administration of benzo- diazepine antagonists, which induce a relatively complete spectrum of anxiety and panic symp- toms. This model, when seen in the light of the wealth of data from in vitro studies, should lead to a comprehensive model of anxiety and panic. Another drug that may provide an interesting model for research in psychopathology is phen— cyclidine (PCP). The close resemblance between PCP-induced psychosis and schizophrenia has recieved frequent comment. In fact, many first ''admissions for schizophrenia have turned out to involve PCP toxicity. PCP has pronounced ef- fects in rodents but, curiously, little stimulatory effect in monkeys. The drug produces a complex array of biochemical effects in animals, and its behavioral effects have been variously attributed to dopaminergic, GABAergic, cholinergic, and serotonergic properties. PCP undoubtedly has some amphetamine-like properties, but its be- havioral actions cannot be attributed to a simple dopaminergic effect. The close resemblance be- tween PCP-induced psychosis and schizophrenia suggests that this drug may provide a useful model. The foregoing outlines the conceptual basis of research using animal models of psychopathology found useful in work in psychopharmacology. A model based on a hypothesis gradually develops and becomes mature when a clinical interface, an in vitro interface, and certain common neuro— biological mechanisms are established. When the model becomes mature, it contributes to a fur— ther understanding of the human psychopathol- ogy not only by the data that support the model but also from the exceptions to the model. The exceptions to the mature model modify the hy- pothesis on which the model is based and, by extension the human psychopathology models. As can be seen, each animal model has limi- tations and restricted applications. Nonetheless, the key temporal aspects of chronic perturba- tion, adaptative mechanisms including altera-— tions in sensitivity, emergent properties, and changes in state in all neuropsychobiological intervention can be studied only in the context of living animal models. Psychiatric disorders and their treatment are long-term, time-— dependent phenomena, with issues of stability and change requiring an intact preparation in which behavior and neurobiology can be studied over long time periods. The major pitfall in a model rests with its re— strictive effects (one behavioral analogy, one transmitter), with a danger that the real clinical phenomenology no longer serves as an inter- active and corrective source. Multiple influ— ences—-e.g., the roles of lactate, NE, endogenous benzodiazepine receptor abnormalities, and conflict—-may be simultaneously present in pro- ducing human anxiety. Many of the current animal models of human psychopathology need to be improved. Perhaps the most apt model is one in which animals are put in a conflict situation that mimics the hypothesized precipitants of human anxiety. Al-— though the amphetamine-or apomorphine- induced stereotypies have resemblances to some of the behaviors associated with schizophrenia, they tell us nothing about the fundamental defect of disturbed thinking. Further, they have locked us into a system whereby we can develop new chemicals that are old drugs. Thus, the model used to screen for new therapeutic com- 59 pounds determines the kind of compounds one will deem to be useful. The various models used for depression are equally inadequate and again tend to produce duplicative drugs. No adequate models are available for mania. Only during the past year or so has any interest developed in producing an animal model of senile dementia. Thus, much work needs to be done to refine present animal models and to produce new ones. Geriatric Psychopharmacology A number of research programs to define neu- rochemical/neuropathological factors responsible for the behavioral deterioration seen in senile dementia are currently under way. One of the more promising leads being pursued emerges from the observation that certain Alzheimer's patients exhibit degeneration of large cholin- ergic neurons in the basal forebrain. These neu- rons, located in the nucleus basalis of Meynert, provide the primary cholinergic input to the cortex and may be functionally responsible for losses of function observed in these patients. Work in this area should continue to try to study the extent to which this degeneration charac- terizes patients with Alzheimer's disease, whether other diseases exhibit similar degenera-— tion in the brain region, and whether other areas or neurochemical systems show similar types of degeneration in Alzheimer's patients. Drugs with known mechanisms of action might provide use- ful tools for some of this work. Efforts to iden- tify effective pharmacological treatment for the clinical symptoms (especially cognitive loss) are now increasing. Notable agents of current in- terest include cholinergics peptides, and mor- phine derivatives. A major problem in geriatric psychopharma-— cology has been the lack of acceptable animal models. Although aged animals have been useful for drug testing, aged animals do not possess the neuropathological changes or the complete pro- file of neurochemical changes observed in the brains of Alzheimer's patients. Attempts to de- velop rational animal models must be encour- aged, and some attention should be given to the use of nonhuman primates. The types of behav- ioral deficits observed in dementia patients, which are most typically associated with higher order cognitive tasks, are often studied more effectively, or with paradigms more closely re- lated to human tasks, by using nonhuman pri- mates. Further, the neurochemical/morpholog-— ical systems mediating these complex behaviors are more clearly homologous in nonhuman pri- mates. The clearly defined nucleus basalis in both human and nonhuman primate brains (as contrasted with a more diffuse, ill-defined structure in rodents) emphasizes further the im- portance of using primates for experimental studies of the nervous system. ''A number of treatment approaches based on the cholinergic hypothesis have been tried but, in general, have been without significant beneficial effects. Interpretations of these negative results might include: (i) the cholinergic hypothesis of memory in general, or of Alzheimer's disease in particular, is wrong; (ii) degeneration of other systems in Alzheimer's disease, in addition to degeneration of the cholinergic system, is im- portant; and (iii) the cholinergic drugs that have been tried—e.g., physostigmine, arecoline, cho- line, lecithin, etc.-are unsatisfactory for one reason or another (duration too short, side ef- fects too strong, or central muscarinic agonist effects too weak). Whichever of these hypo- theses, or any other, are ultimately shown to be correct, future progress will depend on a broad- based research program at the cellular, pharma- cological, behavioral, and clinical levels on cho- linergic mechanisms in general and memory/cog-— nitive/learning mechanisms in particular. Behavioral Toxicology Toxicity testing of medically important drugs has a central place in behavioral pharmacology. Indeed, the first paper to use the term 'psycho- pharmacology' dealt with the effects of a series of antipyretics and analgesics on auditory thresholds in humans. Moreover, the concept of behavioral toxicity was a central one at the on- set of the modern era of psychopharmacology. It is, however, true that only in more recent times has the field of toxicology attempted to be concerned with the neuro- and behavioral toxi- cology of agents found in the workplace and elsewhere in the environment. It is now generally accepted that neurotoxic effects may frequently represent the earliest adverse effects of occupational and environ— mental exposures, even to compounds not com— monly considered neurotoxic. In many cases, the earliest detectable clinical signs of intoxication involve changes in mood, affect, and mental processing. The well-demonstrated vulnerability of the CNS to changes in many types of periph— eral events—such as nutrition, hormonal status, and peripheral organ metabolism—suggests that the CNS may _ respond to _ nonneurotoxic compounds at levels below those associated with functional damage to the target organs. The case of the acquired or chemical porphyrias illus— trates this. The current state of knowledge in behavioral neurotoxicology is uneven. Certain areas, or sub- stances, have received considerable attention— the effects of inorganic lead on the developing nervous system comprise one well-studied sub- ject, as does the production of tardive dyskinesia by antipsychotic drugs. Lead has been investi- gated by basic neurophysiologists, using sophis— ticated techniques of electrophysiology, bio- 60 chemistry, and quantitative morphology; lead also has been studied in several comprehensive clinical investigations using electroencephalo- graphy, performance tasks, and psychometric measures. At present, it is possible to describe the behavioral toxicity of most classic drugs used in psychiatry. Newer drugs are not as com- pletely described. In fact, the field is lacking detailed analysis of the behavioral and neuro- toxicology of clinically important new drugs. Solvents, which are ubiquitous neuroactive sub- stances, are also understudied by both basic and clinical researchers. The development of basic research in behav— ioral toxicology is hampered by continuing dif- ficulties in determining neurotoxic effects in humans: the lack of epidemiological data on either baseline control populations or on selected exposed groups, methods for measuring premor-— bid effects, availability of human material for chemical or other assay, appreciation of stand- ard confounding variables (analogous to the role of smoking in determining cancer risk), widely accepted animal models, hypotheses for predic-— tion across species and over dose ranges (similar to the standard risk assessment approaches used for chemical carcinogens), and rapid assay systems. Neuro- and behavioral toxicology are growing in importance. In occupational medicine, atten- tion is increasingly paid to the neurotoxic ef- fects. In regulation, the testing for neurotoxic effects will increasingly be required. Involve— ment of basic neurosciences in neurotoxicology represents a significant opportunity for growth and the attainment of new knowledge applicable to the basic understanding of CNS function and behavior. Among the areas that can benefit basic be-— havioral toxicology research are: ® Developing in vitro test systems, involving cell culture or acute preparations of neural tissue (synaptosomes, neuromuscular junc-— tion, synaptic membranes, brain slices, neuroblastoma, etc.). The rapid systems for studying receptor function in vitro have not been widely applied in neurotoxicology, al- though they have been used with success in investigative neuropharmacology. e Estimating the functional implications of specific neurobiological phenomena. For instance, what is the normal range of re- gional neurochemical parameters? At what percentage of this range has a threshold implying functionally significant change occurred? ®@ Defining critical populations or especially sensitive groups for the effects of neuro- toxins. It has been assumed that the devel- oping brain has the greatest susceptibility to ''exogenous influences, such as drugs and heavy metals. But the aging brain may also represent an organ of special sensitivity, because of depleted functional reserves, or because age-related rates of loss of specific neuronal populations may be speeded by toxin exposure, producing focally limited premature aging. One sex may be more susceptible to neurotoxic effects of specific agents, possibly because of the sexual di- morphism of the brain and possibly because of differences in peripheral metabolism. As sexual and genetic polymorphisms in brain structure and function are defined, these will be highly relevant to experimental and clinical neurotoxicology. Defining a battery of neurobehavioral tests to discern toxicity is imperative. Behavioral tests proposed to date are intended to be broad, covering aspects of operant behavior, sensor perception, and neuromuscular function. Recommendations for the Future Explicit examples of research needs and opportunities are chapter. Summarized below are those of the panels' recommendations that deal specifically with behavioral analysis of psychoactive drugs. replete throughout this @ Elucidation of the behavioral and neural mechanisms underlying rate dependency and the stimulus properties of drugs e@ Elucidation of the behavioral and neural mechanisms underlying the apparent need for chronic drug treatment before thera— 61 peutic effects emerge Studies on drug effects on social behaviors Studies on the behavioral effects of peptides Chronic in situ studies on neurochemically identifiable neurons in order to discover the interactions of these cells with their target neurons Elucidation of the relationship of changes in receptors and other biological parameters to clinically relevant phenomena (e.g., thera- peutic response) Determination of the signals and mech- anisms that control receptor turnover and the manner in which turnover is altered by drugs, hormones, and disease Design and study of receptor agonists, an- tagonists, and specific neurotoxins to modi- fy the biosynthesis, metabolism, and func- tion of neurotransmitters/neuromodulators Studies of the variations in genomic expression in response to administration of physiological agents and drugs Research on the mechanism(s) of action of psychoactive and therapeutic drugs Development of appropriate animal models of mental disease Studies of the behavioral toxicology of psychotherapeutic agents, in particular, studies of the postnatal effects of drugs taken during pregnancy or administered during development ''CHAPTER V THE NEURAL BASIS OF PSYCHOPATHOLOGY Progress in preventing and treating mental illness depends on an increased understanding of the mechanisms underlying brain function. Al- though knowledge of human psychopathology is the ultimate goal, the majority of experimental approaches available for analyzing brain function cannot be used in human subjects. Fortunately, the nervous systems of other species are known to have many anatomical, biochemical, and physiological features in common with the hu- man nervous system. Thus, the study of neural mechanisms of abnormal behavior in animals is relevant to an understanding of the human con- dition. Knowledge gained about the neural processes involved in the regulation, integration, and development of appetitive, reproductive, and social behaviors in experimental animals has potential bearing on these phenomena in humans. Although the pathophysiology of the major mental illnesses remains largely undefined, sub- stantial progress has been made in understanding the mechanisms of action of the drugs used to treat these disorders. The greatest depth of knowledge lies in the area of synaptic neuro- transmission, although the translation of these alterations at the synaptic level to the level of clinical symptoms and syndromes still eludes us. Understanding at this level requires additional basic information about the functional units of the nervous system that mediate these symp- toms. These neuronal pathways and circuits, and their respective transmitters and connections, are definable with rapidly developing new tech- nologies. Indeed, new technologies will now permit de— finitive tests of many current hypotheses of mental illness and will lead to the development of more refined hypotheses based on new data. Computer-based electrophysiological and im-— aging techniques now allow exploration of transmitter system interactions as well as other complex relationships in the brain. Further, a direct approach to the study of genetic abnor- malities in the mentally ill has been made pos- sible by the advent of recombinant deoxyri- bonucleic acid (DNA) technology. There are three major areas of application of recombinant DNA methodology: (i) the use of complementary DNA (cDNA) probes for the analysis and under-— standing of neurotransmitter systems in the hu-— man brain, (ii) the use of genetic engineering methods for the development of biologically 62 active molecules for use in treatment of mental illness, and (iii) the use of recombinant DNA probes for the analysis of inheritance patterns of mental illness in families. Areas (i) and (ii) are discussed in chapter II, and (iii) is briefly sur-— veyed here. Recombinant DNA and Family Mental TiIness Genetic Linkage. The use of recombinant DNA methods in genetic linkage studies provides the opportunity to study issues that cannot be ad- dressed by other methods. In particular, these techniques impart the ability to provide an es— sentially unlimited array of genetic markers that can, in principle, cover any region of the human genome to the density required for any genetic linkage study. The rate of progress in this area has been exponential over the past several years. As of this writing, the number of such markers exceeds 100. Among the direct applications of this approach is the eventual identification and isolation of single genes shown to play a central etiologic role in a particular disorder, such as the gene responsible for Huntington's disease (HD). The same methods can be extended to analyze more complex situations in which two or more genes are interacting to produce a specific clinical condition, or to situations in which a single gene makes an important contribution to predispose an individual to develop a specific form of mental illness. An explanation of the key features of this technology follows. The strategy for mapping genes coding for neurogenetic disorders depends on collecting DNA samples from pedigrees in which a gene such as the HD gene is segregating. The poly— morphic segregation of DNA probes from every chromosome can then be tested to determine the pattern of inheritance of that DNA polymor- phism in the pedigree. For all marker loci, ex- cept those linked to the HD gene, the pattern of inheritance will be random in relation to the HD gene. Only markers linked to the HD locus gene will show a pattern of inheritance in the pedi- gree that correlates with the HD gene. The location of a gene relative to a set of known marker genes can be inferred using ge- netic linkage techniques. This approach depends on DNA sequences located close to each other on ''a chromosome remaining physically associated during meiosis. In contrast, DNA sequences lo- cated far apart on the same chromosome or DNA sequences located on different chromosomes are less likely to be inherited together. All genetic markers—whether assayed by a visible phenotype, a biochemical phenotype, or any other method—-have the same physical basis: a difference in the sequence of nucleotides of the DNA between each pair of alleles. The dif- ference in base sequence may be a single nu- cleotide, as is the case for sickle cell anemia in humans, or it may involve the presence or ab-— sence of a significant sequence of DNA that may include tens, hundreds, or thousands of nucleo— tides. The nucleotide sequence difference be-— tween the two alleles can be determined without the actual expression of a gene as a visible or biochemical phenotype. Four methods are available for this purpose: (i) the use of re— striction endonucleases to split DNA molecules at specific base sequences; (ii) the separation of DNA fragments by molecular weight using agarose gel electrophoresis; (iii) the isolation, purification, and amplification of a single frag— ment of the human genome by recombination of the DNA fragment with a bacterial DNA vector, which permits replication of the human DNA segment in a bacterial cell; and (iv) the use of the cloned and amplified human DNA segment as a radioactive probe to bind specifically to other copies of that DNA sequence and thereby to identify the positions of these sequences in an agarose gel by the presence of bound radioac- tivity. Taken together, these methods present a powerful approach for detecting variation in DNA sequences at specific chromosomal sites among individuals in the human population. The techniques of molecular cloning make virtually every base sequence in the human genome available for use as a genetic marker. The application of genetic linkage methods to mental illness is essential during the next dec-— ade. These techniques, in combination with recombinant DNA methodology, will accomplish two important goals in this field. First, it will be possible to develop a much clearer understanding of how many genes are actually involved in the predisposition to a particular form of mental illness. This information will be of particular value for diagnosis and genetic counseling. More profound is the potential to isolate and char-— acterize genes suspected of being significant etiological factors in mental illness. The physical isolation of such genes could lead to an under— standing of the physical basis of disease and de— velopment of direct treatment approaches. The highest priority in this area at present is to initiate detailed family studies to identify family groups in which linkage studies can be carried out. The complexities of the inheritance pattern of predisposition to mental illness make it es— sential that large, well—characterized pedigrees 63 be used for genetic linkage studies. Imaging Techniques. Three significant devel- opments have moved neuroscientists close to the capability of safely and routinely acquiring re- gional in vivo biochemical information in the living human brain. First, the appearance within the medical environment of apparatus for nuclear bombardment, such as cyclotrons and linear accelerators, coupled with ingenious techniques for rapid synthesis of radiopharma-— ceuticals, has provided many substances suitable for in vivo regional hemodynamic, metabolic, and pharmacological studies. Second, the parallel development of appropriate mathematical mod- els has provided the basis for practical algo— rithms that allow parameters of biochemical and physiological significance to be estimated from the data. Finally, the development of detection systems employing the concept of positron emis— sion tomography (PET) permits safe monitoring of the fate of these radiopharmaceuticals in vivo in a truly regional and quantitative manner. Emission tomography is a visualization tech-— nique in nuclear medicine that yields an image of the distribution of a previously administered radionuclide in any desired section of the body. Positron emission tomography utilizes the unique properties of the annihilation radiation gener-— ated when positrons are absorbed in matter. In PET, an image is reconstructed from the radio-— active counting data that is an accurate and quantitative representation of the spatial dis— tribution of radionuclide in the chosen section. This approach is analogous to quantitative tissue autoradiography but has the added advantage of allowing in vivo studies in human _ subjects. Especially valuable to the study of psychiatric disease is the potential for PET to provide quantitative in vivo measurements of local drug action, receptor pharmacology, and neuro— transmitter metabolism. Because of the impor-— tance attached to disorders of neurotransmitter and receptor function in the pathophysiology and treatment of psychiatric disease, studies in this area should be especially important. Recent technical advances in the area of brain-imaging techniques have suggested alter— native ways of obtaining quantitative regional measurements of biochemistry and metabolism. This is especially true for nuclear magnetic resonance (NMR). Brain imaging with NMR now appears to offer significant improvements in resolution and contrast over conventional X-ray computed tomography. While the enthusiasm for NMR imaging of the human brain can be appre- ciated by anyone viewing the images, it is not yet clear whether NMR will also provide sig- nificant new information on regional brain bio-— chemistry and metabolism equivalent to that now offered by PET. In addition to NMR, a num-— ber of investigators are stressing the potential ''utility of single photon emission tomography as a less costly and labor-intensive alternative to PET. Although single photon tomography is cheaper and less demanding in terms of skilled personnel, it is less accurate in producing truly quantitative in vivo images, and it lacks almost entirely the broad range of significant radio- pharmaceuticals available with PET. Although significant technical advances have been made in the area of quantitative brain im- aging with PET, the future use of this technique in human neurobiology is critically dependent on parallel investigations in the areas of metabo- lism, biochemistry, pharmacology, and behavior. Future work will benefit from collaboration among neuroscientists, neurochemists, and those familiar with imaging techniques. Special at- tention must be devoted to the training of com-— petent investigators with an interest in mental illness. In addition, there is an obvious need to see innovative approaches to the synthesis of PET radiopharmaceuticals with short-lived, positron-emitting radionuclides, possibly using techniques of molecular genetics, coupled with the development of appropriate mathematical models. Finally, large banks of quantitative im- age data will accumulate in laboratories using PET. Strategies for the utilization of these data by outside investigators should be formulated. We are currently on the threshold of using PET to make a variety of in vivo biochemical, meta— bolic, pharmacological, and hemodynamic meas— ‘urements of the human brain. These studies should provide new and unique information on important clinical and theoretical issues in hu- man neurobiology. In this regard, it is useful to note that brain function consists of complex interactions among biochemical and physiolog- ical processes heterogeneously distributed throughout the organ. These processes are dy- namic in accord with the functional activity of the brain: In the normal brain, it is now clear, metabolism and blood flow are altered in dis— crete regions as a result of highly specific functional activity. These observations provide the basis for a de- tailed exploration of the relationships of various structures within the brain to specific motor, sensory, and cognitive tasks performed by the human central nervous system (CNS). Such studies should provide a better understanding of the organization of motor activity and also il- luminate the manner in which the human brain processes information, stores data, and organizes such higher order functions as language and memory. These data should be helpful in fur- thering our understanding of the functioning of the human CNS and should prove invaluable in our efforts to understand how disorders in human mood and cognition reflect altered local brain function. The foregoing discussion of new technologies suggests that what is needed now is a broader, 64 long-range view of research in mental disorders. Such a perspective will facilitate the integration of new technological and substantive develop- ments in neurotransmission, anatomy, physiol- ogy, and genetics into studies of the neural mechanisms of discrete behaviors important in mental illness. Along with striving for greater integration, research studies should continue in each area of basic biological and biobehavioral research surveyed. Specifically, there should be a continuation of basic studies to identify and characterize transmitter-specific neuronal systems in the brain. A critical mass of information, method- ologies, and testable hypotheses has accumulated with this data base and now allows the study of synaptic neurochemical alterations in brains ob- tained from patients with well—characterized psychiatric disorders. Direct neurochemical analyses of the brains of deceased psychiatric patients should permit the identification of af- fected neurotransmitter systems that may be etiologically involved in these disorders. For success, the approach outlined here will require expertise in psychiatric diagnosis, collaboration with established brain banks, and the active in-— volvement of basic scientists with mastery of current neurochemical and neuroanatomical methods. Those psychiatric disorders with the clearest genetics, neuropathology, or stereotypical de-— fects would appear most promising; these include the affective disorders, schizophrenia, Down's syndrome, infantile autism, and Tourette's syn- drome. Ultimately, anxiety disorders and related conditions will require examination. Identified abnormalities can lead to neurobiologically based animal models. Such models, in turn, can help characterize the role of affected neuronal sys-— tems and their interaction with psychotropic drugs, as is currently being done with Parkinson's disease, HD, and Alzheimer's dementia. Thus, clinical findings provide a focus for investiga— tions by basic neuroscientists using experimental animals, which could lead to more effective clinical exploration by means of the newly de- veloped genetic, mneurophysiologic, neuropsy— chologic, and imaging techniques. Evoked Brain Potentials. The advent of com- puters and the development of sophisticated mathematical techniques make it possible to obtain objective, reliable quantitative neuro— physiological data on CNS function. The quan— titative measurement of salient features extracted from the evoked potential (EP) and event-related brain potential (ERP) recordings reflects various aspects of brain function related to sensory, perceptual, and cognitive processes as well as the structural and functional integrity of different neuroanatomical systems. These EP techniques occupy the interface between cellular neurobiology and the behavioral or cognitive ''sciences. Indeed, a great variety of information about brain function can be obtained by analyz-— ing the evoked electrical activity that can be recorded from the intact human scalp. Sensory evoked potentials provide objective measures of the functional state of the afferent sensory pathway, including the receptor, primary afferent neuron, ascending sensory pathway, and within the CNS, specific sensory cortex, and nonspecific cortex. While knowledge of the par- ticular generators for each of the components of the evoked potential is incomplete, sufficient information is available to enable the clinician to utilize EPs in patients with disorders of sen— sory and neural functions. In recent years, these sensory evoked potential methods have been de- veloped and utilized in several clinical areas with notable success. Surveyed here are various potentials that can be evoked by sensory stimulation. Auditory brain stem potentials can be clas- sified as abnormal, depending on changes in morphology, voltage, and conduction velocity. For instance, a focal lesion within the brain stem will be accompanied by a latency change re-— stricted to a particular component or compo-— nents. An alteration of central conduction time is typically found in patients with multiple sclerosis. Deficits in conduction velocity, indi- cating the presence of brain stem aberrations, have also been reported in many alcoholic pa- tients. Studies of brain stem auditory evoked potentials in schizophrenic patients have all re- ported normal findings. Pattern-reversal stimulation has emerged as a useful and relatively simple method for visual evoked potential studies. Accurate identification of visual disturbances has become a reality with the use of this method. Abnormalities of the visual evoked potential are not specific for a given disease but rather indicate disturbances of function somewhere in the visual pathways. With respect to the somatosensory evoked potential, there is evidence that activity in ascending portions of the somatosensory pathway is mani-— fested by potentials from the medulla, brain stem, and thalamus as detected by far-field po- tential techniques. This technique is used to as-— sess nerve conduction velocities (as in peripheral neuropathy), to define spinal cord function in spinal trauma, to localize lesions of the somato— sensory pathways, and to assess disorders of pain and temperature. Somatosensory evoked poten— tials are quite stable and consistent within the normal population and vary systematically with the physical parameters of the stimulus. In contrast, ERPs associated with perceptual phenomena are far more sensitive to variables of subjective state. The perceptual experience of a stimulus event depends not only on the event's physical characteristics but also on the mo- mentary state of the observer—level of alertness, attention, memories of relevant past 65 experiences, expectations, and motivation. The fact that ERPs are associated with these kinds of variables (in a highly complex, interactive fashion) spurs efforts to identify signs or codes of perceptual processes at the level of these field potentials. At present, the scalp-recorded ERPs provide the major window to the neuro- physiological transactions of the human brain as it processes information on a millisecond-to- millisecond basis. ERP studies have been used to clarify attentional mechanisms on both physio-— logical and psychological levels. The timing and localization of specific ERP components that vary with selective attention may permit in- ferences about the levels of the nervous system where stimulus selections take place. Disturbances of selective attention underlie several types of clinical disorders, including the affective disorders, schizophrenia, minimal brain dysfunction, attentional deficit disorder, alco- holism, dementia, and learning disability. The attention-related ERPs have obvious applica— tions for the characterization of such attentional deficiencies, for the diagnosis of different syn- dromes, and for the evaluation of therapeutic regimens. For example, deficits in the N1 com-— ponent of the ERP have been reported in chil- dren with minimal brain dysfunction, in hyper- active children, and in children with learning disability. Another example concerns. the assertion that a major aspect of the psycho- pathology of schizophrenia is an aberration of attention. Recent examination with ERP tech- niques suggest that the Nl component of the ERPs of schizophrenic patients are indeed markedly abnormal. A number of endogenous components of the ERP have been established and examined, in- cluding the P3, a robust endogenous component that is reliably recorded in association with task-relevant, rare stimuli. The P3 component recently has been associated with biologically significant stimuli. Intracranial studies in hu- mans have demonstrated that the hippocampus is involved in the production of this component. The P3's functional significance is sufficiently well established to make it a strong candidate for study in psychiatric populations. A signifi- cantly reduced P3 component in patients with schizophrenia has been observed consistently and reliably across many laboratories. The expectations that motivate these studies hold that it will be possible to analyze and characterize the qualitative and quantitative relations between the stimuli used and the electrical responses obtained. In so doing, it will be possible to fractionate the complex neural system into functionally and operationally dis- tinct subunits. Study of these potentials should provide valuable information concerning the lo- cus and patterning of neural activity involved in processing sensory input, arriving at some deci- sion, and executing appropriate behavior. Exam- ''ination of these electrical phenomena in well- defined psychiatric populations may shed light on psychopathology and neuropathology. Localiza— tion of the sources of EP components generated within the brain is one of the most fundamental issues in this area. In addition to source genera-— tors, the neurochemical systems that subserve these electrical phenomena should also be exam-— ined in depth. Brain Function in Psychopathology The past decade has witnessed remarkable advances in understanding of the brain's func— tional organization at the level of synaptic neu-— rochemical mechanisms and the selective in— teractions of clinically effective psychotropic drugs with these processes. Through the appli- cation of histochemical and immunocyto- chemical techniques, a transmitter—specific neuroanatomy is now well developed, especially with regard to the cortical, limbic, and reticular core systems relevant to psychiatric disorders. Through the application of ligand-binding auto-— radiographic techniques, an integrated picture of synaptic processes of the brain is being elabo- rated and correlated with psychotropic drug ef- fects. These advances, coupled with a rigorous approach to human psychopathology, have led to the development of a number of neurobiologi- cally oriented hypotheses about the pathophys— iology of the major psychiatric disorders. Concurrent with advances in the understanding of brain transmitter processes, clinical and basic neuroscientists are beginning to address synaptic neurochemical dysfunction in post mortem brain material from patients with psychiatric dis— orders. HD and, more recently, Alzheimer's de— mentia have been the focus of interest because of the opportunity for correlating human psy-— chopathology with neurochemical and histologic alterations. Studies in this area have provided a detailed map of the affected systems in HD and have radically altered the conceptualization of the pathophysiology of senile dementia of the Alzheimer's type, the most prevalent and de- bilitating form of cognitive deterioration. Basic studies have established that antipsy— chotic drugs and certain psychotomimetic drugs (e.g., amphetamine) interfere with catechola— minergic neurotransmission in the brain, espe— cially dopaminergic synaptic transmission. Other work has revealed that the monoaminergic neu-— rons apparently possess receptors on their own soma and/or dendrites that are specific for their own neurotransmitter. It is now thought that a number of psychotomimetic drugs act either directly or indirectly on such receptors, and that these sites of action are significant for the con— trol of presynaptic neuronal activity and for neurotransmitter biosynthesis. This notion has been extended to include receptors for psycho- 66 tropic drugs that represent unconventional and novel sites of action—-e.g., on the axon terminals of neurons that innervate catecholaminergic neurons or their targets. Neurotoxins. A new and potentially useful method of altering brain development in labo- ratory animals is the prenatal administration of mitosis inhibitors such as methylazoxymethanol (MAM) or cytosine arabinoside. MAM is partic- ularly useful, because it appears to be selec— tively toxic to neuroblasts in the final stages of division while sparing fully differentiated cells undergoing proliferative mitosis. Thus, adminis— tration of MAM during histogenesis of the neu- ronal population of interest may provide a means of mimicking developmental insult to the brain. This approach may become of particular interest in view of recent associations between mental disorders and brain abnormalities found by com- puterized tomography scans, PET scans, brain electrical activity mapping, cerebral blood flow, and post mortem studies. Early investigations of the effects of MAM-— induced developmental arrest on learning in animals failed to show any gross behavioral deficits in MAM-treated animals. These animals were similar to controls in most measures of learning, although some substantial deficits were seen in maze-learning; MAM-treated animals were also found to be hyperactive. More re- cently, it has been shown that prenatal treat-— ment with MAM causes a marked increase in the dopaminergic, serotonergic, and noradrenergic innervation of the adult neocortex without sim-— ilar effects on cholinergic and gamma amino-— butyric acid (GABA) activity. These findings provide further support for the idea that pre- natal administration of MAM may constitute a method of altering brain structure comparable to the kinds of damage postulated to be responsible for some human psychiatric disorders. Neural Transplants. The technique of grafting neuronal and endocrine tissue into the brain has recently aroused a great deal of interest. The brain has for many years been known to be a privileged site, in that it will accept tissue grafts that would be rejected if transplanted to other sites. It has been demonstrated that fetal donor tissue grafted to the brain of a neonatal animal achieves a substantial degree of inte- gration. This is probably because both the donor tissue and the host brain retain a large degree of developmental plasticity. By contrast, the growth potential of fetal grafts in adult host brains is considerably more limited. Recent studies have shown that brain tissue grafts can have functional consequences. Brain grafting is of clinical interest as a potential method of repairing the effects of functional cell loss. The brain structures involved in psy-— chopathological states are not yet well enough ''understood for the application of this technique in the foreseeable future. The study of brain grafts might be expanded by making technical improvements in transplantation within estab-— lished models (e.g., recent results showing that large areas of the caudate nucleus may be re- innervated by producing many small grafts in the form of dissociated cells) and studying trans— plantation in fetal or prenatal recipient animals. Ultimately, exploitation of the clinical potential of brain grafting will depend on increased un- derstanding of the factors that guide and pro- mote brain development (see chapter I). Developmental Psychobiology For the past several decades, researchers in-— terested in the neural mechanisms of abnormal behavior have utilized experimental stress par-— adigms. These, in general, produce abnormalities of behavior and, presumably, of the underlying physiological systems. Stress models have been widely used in developmental studies as well. Typically, infant animals are exposed to various physical stressors, such as handling, drugs, and nutritional, sensory, or maternal deprivation. In adulthood, the behavioral and physiological status of these animals is studied during stressful situations. The paradigm is intended to produce exaggerated responses in adulthood that will yield unambiguous results. It has produced some interesting, and at times controversial, findings. The small, biochemically abnormal brains of rats raised in relatively impoverished environments and the bizarre social behavior of isolated mon-— keys are two well-known examples from a grow- ing collection. In recent years, developmental psychobiology also has begun to focus on the relationship between early experience and adult behavior. This emphasis on developmental proc-— esses has already opened several promising avenues for studying the pathophysiology of aberrant behavioral states. Developmental Stage. Perhaps the most gen- eral of these new insights is that neural regula— tory systems are organized differently in the young mammal than in the adult. Additionally, there may be two or three transitional stages in the postnatal development of any given system. One example of this receptor staging idea is seen in the observation that administration of halo- peridol to pregnant rats produced offspring who as adults had decrements in dopamine receptors and reduced responsiveness to apomorphine, whereas the same drug given postnatally had the opposite effect. In another recent study, it was found that sectioning the chemoreceptor nerves of the carotid sinus and aorta in infant rats produces a disorder of respiratory regulation characterized by periodic breathing and apnea during rapid eye movement (REM) sleep. This 67 experimental disorder is more severe in the young pup, with death occurring between 12 hours and 10 days following surgery in more than half of the animals. By the age of 3 weeks, how- ever, pups respond like adults, showing little or no respiratory dysregulation and no mortality. Apparently, peripheral chemoreceptor feedback plays a vital role during REM sleep in early in- fancy. These findings may be relevant to the etiology of sudden infant death syndrome. Mother-Infant Interaction. A second insight from developmental psychobiology that is shap- ing new research is the extent to which the mother is the source of physiologically relevant environmental stimulation for the infant mam— mal. It appears that specific aspects of the mother-infant interaction regulate certain of the infant's neurophysiological and biochemical systems. For example, it has been found in rats that the rate and quality of milk provided by the mother regulate autonomic neural organization in the infant. There is also evidence that certain forms of tactile and olfactory stimulation may serve to regulate the behavioral reactivity of the infant by affecting the accumulation of brain norepinephrine and dopamine. Other research provides evidence for similar regulation of brain and cardiac ornithine decarboxylase in rat pups. Several more examples of such processes have recently been reported, and it seems likely that more will be discovered as this work progresses. The implication of this approach is that the mother-infant interaction may be accessible to physiologic analysis in addition to the more traditional approaches that depend on inferred emotional states. The mother-infant interaction approach also provides the basis for a reinter-— pretation of maternal deprivation effects. The devastating effects of prolonged absence of mother-infant interaction on subsequent social development have been studied almost entirely in terms of inferred affective states and the effects on learned behavior. This formulation has been successful in providing a means of under-— standing the acute response to maternal sepa— ration—the vocalization, motor agitation, and associated neurohormonal activation. But the slower developing effects of maternal depriva— tion have not been satisfactorily explained. Be— haviors such as sterotyped acts, unusual postures, self—stimulation, rocking, motor retardation, in- hibition of play, and outbursts of aggression, as well as physiological changes including lowered body temperature, sleep-state disorganization, lowered heart rate, and increased susceptibility to intercurrent diseases, all develop over days or weeks after separation. Although some of these problems eventually correct themselves without return of the mother, many do not. If one takes into account the possibility that separation of the infant from its mother results in the alteration of a whole ''set of regulatory mechanisms, many of these slower developing effects may come to be ex-— plained on a physiological basis. This pattern of changing physiological regulation has, in fact, been found to occur in the 2—week-—old rat pup over 5 or 6 days following maternal separation; it is quite distinct from the classical acute sep— aration distress response that occurs in the first minutes after separation. The concept of the mother as a regulator in early development suggests various research strategies. One would be to seek adequate sub- stitutes for specific aspects of the mother-— infant interaction and look for relevant neural systems. In the maternally deprived Rhesus monkey, for example, it has been found that stereotyped rocking can be completely prevented by providing vestibular stimulation. This is or- dinarily given to the infant as it clings to its active mother. Another strategy would interrupt specific sensory inputs of a maternally deprived infant and then return the infant to its mother. Resultant changes in mother-—infant behavior or infant physiology could then be observed and documented. Experimental olfactory deprivation, for example, has been found to produce a number of changes in mother-infant regulatory systems. Such interference may also change the mother's behavior toward her infant, which in itself could affect the same or different systems, compli-— cating interpretation of the results. A third strategy might proceed from the possibility that certain kinds of interaction have unique long- term developmental effects. If this were so, the quality of parenting could be analyzed in the same way that the effects of quantitative re-— ductions of deprivation have been studied. Genetic Approaches. An important approach to be mentioned is selective breeding and cross— fostering. This approach has generally been used as a control procedure in genetic studies to establish that two substrains, bred for dif- ferent behavioral characteristics, are not being produced as a result of differences in the post- natal environments. Yet, there have been several studies reported where the control group was affected. In one study, for example, pups born to normotensive Sprague-Dawley rats grew up hy-— pertensive when cross-fostered at birth by mothers of the spontaneously hypertensive Wistar strain. In another study, two laboratory subpopulations of Wistar rats, one bred to be mouse-killers and the other bred not to kill mice, were found on cross-fostering at birth to grow up with the behavioral characteristics of the postnatal mother. In neither of these model situations is it known which aspect of the inter— action was responsible for the altered behaviors. However, the cross-fostering technique allows the generation and identification of qualitative differences in the postnatal maternal input that have known long-term effects on the biology 68 and/or behavior of the offspring. Further analyses of these effects should yield new in- sights into the processes underlying the devel— opment of behavioral disorders. Among the most promising future directions are those based on new methods and concepts for understanding maternal-fetal and maternal-— infant interrelations. Techniques now exist for dissecting the behavioral and developmental features of the young organism. There is con-— siderable potential for the development of re- search in the study of prenatal development, fetal behavior, and early postnatal development. Possible specific areas include prenatal sexual differentiation, prenatal behavioral develop-— ment, transmission of information through pre— natal experience, behavioral effects of prenatal alterations of brain functioning, and in vivo ap— plication of substances that are thought to in— fluence prenatal development of the brain, e.g., hormones and growth factors. Learning and Motivation. Another area of current research involves the study of early learning and motivation. Here, too, old ideas have had to be altered in the past few years, and new approaches seem likely to provide basic knowledge about the formative stages of neural systems underlying disorders of affect and cog- nition. Until recently, learning capabilities and the organization of motivational systems in in- fants were thought to be at least as undeveloped as their sensorimotor performance—for ll practical purposes to be nonexistent. Lack of myelination in the CNS of the infant rat and low levels of brain neurotransmitters prior to 8-10 days of age were interpreted as evidence for the futility of looking for higher integrative func- tion. A few experimenters did look for evidence of learning and motivation in infants but used techniques and situations developed for study of adults. Their negative results supported the traditional assumptions. In the past few years, investigators have begun to analyze the existing motor capabilities of the infant rat. Using ingenious experimental tech— niques based on this knowledge, they have been able to demonstrate relatively sophisticated cognitive and motivational feats in neonatal rats. For example, instrumental learning (and food ingestion) in the absence of the mother has been elicited from 1-day-old rat pups, who are neurophysiologically similar to the early third trimester human fetus. In rat pups only 3 days old, self-stimulation can be elicited from basal forebrain sites, and in 8-day-olds, one-trial avoidance learning can take place and be re- tained for days. The intake of nutrients and water is sensitively regulated by 1-week-old pups in certain situations and can be used in studies of the early states of motivational organization. There is even a period of infantile amnesia, as in the human, and evidence of an ''inhibitory influence exerted by the mother on certain types of learning contingencies. Recent research with human neonates has disclosed their capacity to discriminate their own mother by smell and sound, to perform relatively complex learning: schedules, and to imitate adult facial expressions. Understanding these capabilities, plus the knowledge being gained about the special environmental and stimulus conditions needed to support them, provides the basis for a new area of exploration of the neural bases of cognitive and motivational disorders. The neural mechanisms underlying infant learning and motivational organization may be, relative to adult features, easier to de— lineate. Once understood, they may generate new ideas about how to approach the analysis of adult systems. Sexual Differentiation. Experimental studies on sexual differentiation indicate that the brains of virtually all mammalian species can be per— manently altered by the action of sex hormones during development. These studies have provided behavioral, neuroanatomical, and neurochemical data on which to base studies of human psycho-— sexual development and differentiation. Devel— opmental effects typically occur only during a time-limited critical phase of development and tend to be relatively permanent, although they can be overridden by nonhormonal influences. Some developmental effects are delayed and are behaviorally manifested only in later life. The delayed appearance of these effects may depend on the interactions of sex hormones during puberty or adulthood. Studies of sexual differentiation have focused on those behaviors that are typical for one or the other sex, i.e., aggressive behavior, vocaliz— ation and courtship behavior, mating behavior, etc. It is important to keep in mind that almost all so-called masculine and feminine behavior characteristics are not truly sex—dimorphic. For instance, male behavior has been observed in females of several species of mammals, as have (less commonly) spontaneous displays of female sexual behavior by male mammals. This suggests that hormonal manipulation may serve to in-— crease the frequency of the occurrence of a specific behavior in one sex or the other rather than effect a true reversal of behavior. Varia— bility in the expression of behavior by both sexes points to the important principle that biological factors do not act in a vacuum but interact with other influences. Sufficient evidence now exists to indicate that biological factors make a substantial contribu- tion to psychosexual differentiation in humans. Anatomically and histologically, the subcortical regions of the human brain have a striking sim— ilarity to those of subhuman mammals. The timing of peak androgen production by the fetal testes corresponds closely to the timing of hy— 69 pothalamic differentiation, much as it does in the lower mammal. Also, the effects of human sex hormones on the differentiation of the re— productive tract are the same as seen in most other mammals. That is, independent of genetic sex, if a fetus is exposed to high levels of pre- natal androgen, the differentiation of the ex- ternal genitalia will be masculine; in the absence of androgen, the external genitalia will be feminine. One of the most important aspects of sexual behavior currently being studied is gender iden— tity. This is a specifically human concept; there are no animal analogs. In human psychosexual differentiation, gender identity is critical, since identification with one sex or the other serves as an important ordering principle and influences other actions and reactions. The predominant view is based on early observations of intersex patients and holds that, rather than any single biological factor, social learning is of paramount importance and can be the decisive factor in predicting a person's gender identity. Recent work has demonstrated that human beings may be capable of changing their gender later in life, but only under special and specific circum- stances. The conditions that facilitate such a gender change at this late stage do not appear to be determined by either prenatal or pubertal sex hormones. One consistent sex difference in behavior that appears to be truly sex-dimorphic and may be related to prenatal sex hormones is rough, ag- gressive play behavior. In monkeys, this behavior includes chasing, hitting, wrestling, and biting, all of which are seen more often in males than in females. Similarly, male human children display motor patterns described as pushing, hitting, pulling, chasing, and wrestling. In cross—cultural comparisons, boys are found to engage in more rough—and-tumble play than girls do in a variety of societies. Another salient sex difference involves nur-— turance or parenting behavior. While females are usually observed to be more nurturant than males, the curves overlap considerably. What is the evidence that these behavior differences are influenced by prenatal sex hormones? The orig-— inal human studies focused on females who had been exposed to abnormally high levels of pre— natal androgens due either to maternal intake of a specific type of hormone during pregnancy or, more frequently, to an endocrine abnormality, e.g., congenital adrenal hypoplasia. The behavior of these prenatally androgenized girls differed from that of controls, in that they typically demonstrated a combination of intense, active outdoor play, increased association with male peers, and long-term identification as tomboys. They also showed decreased parenting rehearsal in doll play and baby care and a low interest in the role rehearsal of wife and mother. The characteristic pattern, which is not considered ''abnormal for females in our culture, persisted throughout childhood. More recently, several studies have focused on the long-term effects of progestogen treatment, during pregnancy. The treatment of high-risk pregnancies with such agents was introduced in the 1940s and has been widespread. It was orig— inally suggested that progesterone may protect the developing hypothalamic—pituitary—gonadal system of the fetus against androgen or estrogen or both. According to this theory, progesterone protects the brain from the masculinizing ef- fects of androgen. Thus, abnormally high levels of progesterone may antagonize fetal testicular androgen in males; in females, it may antagonize the smaller amount of androgen produced by the adrenal. In both sexes, the expected effect is greater feminization or demasculinization. Higher Order Integrative Processes Research on the neural basis of higher order integrative activities—-attention, perception, and cognition—currently focuses on the basic mech-— anisms for the coding and representation of in- formation in the nervous system. New methods have made it possible to determine the “birth— days" of neurons that form specific structures, and to follow their migration from germinal zones in the neural plate state through the dif- ferentiation of specific cell classes and to the formation of functional synaptic connections in particular structures. Developmental neuro— biology has elucidated the maturation processes of the cerebellum, spinal cord, and portions of the visual system in embryos and fetuses. This work lays the foundation for understanding the ontogeny of those portions of the nervous system critical for perceptual and cognitive processes as well as sensorimotor behavior. Information Processing. Progress in the area of information processing has been extraordi-— nary. Discoveries about the organization of pri- mary sensory areas, in particular the primary visual cortex, have been among the most sig-— nificant. Cortical cells have been demonstrated to show selectivity for coding specific aspects of the sensory environment, such as orientation of line segments or spatial frequencies. The central representation of these parameters has been shown to be organized as systems of repetitive modules on the cortical surface. Another sig— nificant discovery is that there are multiple representations of the same sensory surface, each area possibly subserving a different mode of analysis of the sensory input. Progress has been made in identifying parallel channels of information processing, e.g., the X and Y cells originating in the retina. Finally, there has been considerable success unraveling the nature of 70 sensorimotor integration in vision and eye move- ments. These several achievements were possible because of technical advances in anatomical tracing techniques, intracellular recording and labeling, neurochemical analyses of brain tissues, chronic single unit studies in behaviorally trained animals, and computer-aided quantitative methods. ; More than three-quarters of the cerebral surface in man can be considered association cortex. The association cortex, like the primary. sensory areas, is columnar in organization; inputs and outputs are channeled and distributed in precisely organized geometric arrangements. The columns may be functional units, for they are active as metabolic modules. Other work reveals marked differences among cortical re- gions in the content and activity of monoamines. The frontal cortex is rich in dopamine, a neuro- transmitter that has been implicated in a variety of psychiatric and neurologic conditions. Deple- tion of dopamine in frontal cortex can produce cognitive deficits. Moreover, selective loss of dopamine in this region occurs in aging primates, as well as in patients with Parkinson's disease. Studies of parietal association cortex have opened up possibilities for studying complex perceptual tasks at the cellular level. The first electrophysiological recording studies of frontal association neurons in awake, behaving monkeys found evidence of functional specificity. Finally, increasingly detailed knowledge about the con- nectivity of forebrain structures has begun to reveal their functional organization. Detailed information—processing models have been developed in the functional areas of reading and listening. These models describe several levels of internal coding and their interconnec— tions. The models provide an impressive ap- proach to the way in which the brain functions during realistic cognitive activities that can be accessed from visual or auditory input channels. A number of important and novel results have been obtained. For example, the meaning of visually presented words can make semantic contacts without the use of internal speech; un— derstanding a word's meaning can be viewed in terms of complex networks, many items of which can be activated by any single word; and such semantic activation may occur outside the sub- ject's conscious awareness. These findings have made the study of normal cognition particularly interesting, in that new light can be cast on in- ternal processes that are not available to report by the subject but are available as a result of detailed experimental analysis. These findings from normal cognition, made possible with the techniques of modern neuroscience, have already begun to have considerable impact on our un-— derstanding of pathological states. The presence of new methods of spatial imaging may provide the basis for linking some of these internal cog- nitive processing components to underlying neu- ''ral systems. Such work has already begun to make some progress in the area of spatial attention. Finally, electrophysiological approaches are being made to questions about higher order in— tegrative processes. Noninvasive recording of ERPs from the intact scalp provides an interface between cognitive psychology and human brain physiology, on the one hand, and human and an- imal studies of the neural basis of behavior, on the other. Using these techniques, progress has been made in analyzing the mechanisms of se- lective attention. Specific ERP components can be elicited by attended channels of information, a capability that has served to clarify the structure and timing of stimulus selection proc-— esses. Experiments using auditory and visual stimuli indicate that different sensory features are processed in parallel, but the selection of simple features precedes more complex stimulus analysis in a hierarchical fashion. In the visual modality, there are striking parallels between the ERP components associated with spatial at— tention in the human and attention-related mod- ulations of neuronal activity in the parietal lobes of the monkey and the human. A series of ERP components has been identified as signs of stimulus evaluation and classification processes. Studies of the timing of these ERPs during de-— cisionmaking and memory search tasks have made it possible to parcel out the separate con-— tributions of sensory and motor processes to task performance. The processing of language stimuli also has been associated with specific ERP signatures on the scalp. Using these methods, the semantic analysis of written and spoken messages can now be undertaken, and the relative contributions of different cortical regions to language processing can be assessed. It is now also possible to in— vestigate the subcortical generators of these potentials by recording the weak magnetic fields produced by ion currents within the CNS during neuronal activity, through the use of the non-— invasive technique of magnetoencephalography (MEG). Measurement of the brain's magnetic field through the use of MEG offers advantages over the method of electroencephalography (EEG) and the possibility of obtaining new in- formation. For example, the EEG records a summated measurement from a wide area, while MEG is capable of recording from a confined region; the EEG is affected by the skull, while MEG is not; the EEG is a differential measure, while MEG is absolute. Because of these char-— acteristics, MEG can be used to study many im— portant electrical events. One extremely im- portant application of MEG would be the study of subcortical generators of ERPs in humans and other species. Higher order integrative processes are par-— ticularly amenable to the following lines of research. 71 @ Studies of normal human perceptual, cog- nitive, linguistic, and attentional processes, utilizing such techniques as activity imaging (NMR, PET) and noninvasive electrophys— iological responses (ERPs, MEG, EEG) to examine structure-function relations as well as ongoing physiological changes in the brain @ Studies incorporating the most productive new concepts in cognitive psychology, e.g., those associated with the processes of se— mantic priming and spatial selective pri- ming, and those dealing with distinctions between semantic and episodic forms of memory @ Studies of patients with various forms of injury and disease of their nervous systems e Studies of animals that combine current anatomical, physiological, and pharmaco- logical techniques with innovative behav-— ioral paradigms to study the structural and functional organization of higher brain centers (somesthetic, auditory, and visual cortices in which the sensory surface has been shown to be repeatedly represented, as well as parietal, temporal, and frontal as— sociation cortices) ® Studies of animals that examine the con- tribution of specific neural structures or pathways for information processing using methods of inactivation (preferably revers— ible), while simultaneously monitoring be— havioral performance or specific neural activity Learning. Studies of the neural basis of learning and memory show that neurons are capable of being modified by experience (see chapter I). Current work seeks strategies through which to relate these findings in isolated neu- ronal systems to learning. The model system approach to the cellular basis of learning, used with simplified invertebrate systems, has gen- erated important findings. Indeed, the progress realized with invertebrate preparations is vali- dating the model system approach itself and is strengthening the assumption that learning can be studied at a cellular/molecular level. More- over, the invertebrate systems that have been used are proving to have a broader behavioral repertoire than initially envisioned, such that they promise to be productive for the study of associative as well as nonassociative forms of learning. The few such systems currently being considered indicate the feasibility of such anal-— yses; nevertheless, it is essential that vertebrate model systems be developed. It appears that the powerful analytic techniques applied to the in- vertebrate systems—e.g., the study of ion channels with voltage and patch clamp tech- ''niques—may be applicable to vertebrate sys- tems as well. The development of vertebrate models is being facilitated by the remarkable advances in methods for studying the connec-— tivity of the vertebrate nervous system in vitro. The model system approach to analysis of the neuronal substrates of learning and memory has been productive in the study of simple forms of learning. Habituation has proved a useful choice, for it exhibits similar behavioral properties and, to the extent analyzed, similar neuronal mech-— anisms, in a range of animals from invertebrates to humans. One of the most difficult conceptual issues in the study of learning is that of learning vs. performance. At the behavioral level, learning is an inferred process; all that can be measured is performance. Thus, when biological variables altering learned behavior are used, their actions should be assessed in relation to aspects of performance, e.g., motor responses, sensory processing, motivation. Many studies utilize classical conditioning, which permits in- dependent assessments of effects on learned response (conditioned response) and on perform-— ance of the response itself (unconditioned re- sponse). Classical conditioning also permits es- tablishment of a causal link between stimulus and response. Little is known about the neuronal substrates of more complex forms of associative learning. Suitable paradigms involve some combination of stimuli, responses, and reinforcements—-some process of pairing analogous to the pairing of the conditioning stimulus and unconditioned stimulus in classical conditioning. In all cases, the critical requirement is some degree of contiguity. At least for simpler forms of learning, some form of reinforcement, appetitive or aversive, also ap— pears necessary. The fundamental goals in the study of neuronal substrates of associative learning and memory are (i) identification, lo— calization, and characterization of the brain systems, regions, and neurons that operate to store and retrieve memory and (ii) analysis of the cellular/neuronal mechanisms underlying memory, storage, and retrieval. A number of paradigms and systems are available to test the generality of newly dis— covered neural mechanisms. Paradigms viewed as most promising are of two general types. The first makes use of well—characterized model systems, where cumulative understanding of anatomy, physiology, biochemistry, and behavior should permit analysis of learning and memory at a cellular or synaptic level. Associative learning appears particularly ripe for this kind of analy— sis. The second general approach endeavors to achieve a global description of learning and memory at the neuropsychological level by de-— termining which brain regions are involved and characterizing their roles. Ultimately, this ap- proach aims at an information-processing or flow diagram description of how and where 74 memory storage is accomplished by the brain. Models of Psychopathology In the neurosciences, one of the major ap- proaches to understanding normal brain function is through the detailed analysis of damaged or disordered brain function. One approach to an understanding of memory, for example, has come from the analysis of defects of memory in am- nesia patients. Memory. On the basis of these kinds of stud- ies, it is now possible to distinguish between two kinds of memory systems: the "knowledge-how" system, which is involved in the acquisition of new skills and procedures, and the "knowledge- that" system, which is involved in the recording of facts or data. Amnesia victims, while losing the ability to establish a record of facts or data, preserve the ability to acquire new motor and cognitive skills. Because it is the integrity of the medial temporal and diencephalic structures that is damaged in amnesia patients, the knowledge-— that system is tied to these structures. Cor-— respondingly, the role of these structures in memory is selective and does not affect the acquisition of knowledge-how, that is, the acqui- sition of new skills and procedures. Thus, the neural changes that underlie this kind of learning may be contained entirely in neural systems out-— side of the medial temporal region. The concept that discrete brain changes are responsible for memory has been the motivating force behind much of the work in both verte- brates and invertebrates. This work is based on the assumption that a functional synaptic change is the place to look and the approach to take in trying to understand how the brain stores new information. The cognitive neuroscientist takes a different view and approaches the problem from a systems perspective. The phenomenon of mem-— ory is recognized as being multidimensional. That is, any piece of information the organism stores is coded by separate, superordinary sys- tems for time, space, and affect, to mention but a few dimensions of human memory. Each of these aspects of stored information has been demonstrated to exist and to contribute to the memory process in an automatic fashion. Recent studies suggest that this automatic process is disrupted by brain damage, as evidenced by striking deficits in the ability to recall recent information. Language. Cognitive neuroscience has ex- plored the language process through numerous studies of split—-brain patients. In brief, the right hemisphere of humans does not normally possess language processes; yet, in a very small number of split-brain patients, language does seem to be present in the isolated right half-brain. This ''allows one to compare the cognitive capacity of those rare isolated right brains possessing lan- guage with that of right hemispheres lacking language. Using this strategy, one can assess what language confers on a neural system that does not normally possess it. Studies to date show that language-competent right brains can— not make simple inferences, cannot do simple math, cannot solve simple spatial relation prob- lems, and cannot generate simple images. In short, it is a half—brain that has some syntactic capacities, a half-brain that can even speak, yet cannot make computations from its language data structures. From this view, it is as if lan— guage in the left hemisphere may be merely re- porting out the final computational products of events going on in other cognitive systems. Neuropathology. Although structural brain lesions have not been found in association with the major psychiatric disorders, analogous symp-— toms are often associated with disorders char- acterized by an observable lesion. These dis— orders may serve as models of certain aspects of the so-called functional disorders. Similarly, animals with known lesions may permit the development of rational animal symptom models. Our current knowledge about schizophrenia, affective disorders, and the other major psy- chiatric illnesses is inadequate to validate ani-— mal models by etiology, neuropathology, or clinical manifestations. Two principles can guide research that may lead to progress in this area. One is to emphasize the relationship between neuropathology and behavior, and the second is to broaden the scope of projects with recognized potential relevance to mental disease. Broadening areas of interest should include disorders where one can identify a specific neu- ropathology, and in which the emotional or psy- chopathological manifestations can be docu— mented through careful clinical observation. Examples of these would be focal brain injury from stroke, traumatic brain injury, Parkinson's disease, HD, mental retardation, dementias, and delirium. A further example is depressive dis- orders that occur in some stroke patients. By studying patients who have defined lesions, neu— ropathological-psychopathological correlations have been established. Stroke lesions found closest to the frontal pole on X-ray computer tomography scans are strongly associated with an increasing severity of depressive symptom- atology. These findings implicated left frontal brain regions in the development of depression and define an anatomical—behavioral correlation. By replicating this lesion in laboratory animals, it has been possible to show a linear relationship between the degree of change in norepinephrine concentration in the cortex and the proximity of the lesion to the frontal pole. This animal model of stroke can then be documented in terms of its neuropathology, and the behavioral manifesta— 73 tions of discrete focal brain lesions studied. A similar approach is possible with numerous other clinical conditions. Animal models can then be used to investigate mechanistic elements of the brain-behavior relationships; specific neurotransmitter pathways can be investigated, and their importance in eliciting particular be-— haviors can be determined. This approach has already been very useful in the investigation of the role of the nucleus basalis of Meynert in de- mentia. It is important that the breadth of con— ditions that are examined be increased. It is equally important that the specific entities chosen have either a definable neuroanatomy or neuropathology that will allow clear investiga- tion of the brain—-behavior relationships. Although such conditions as dementia, delir— ium, stroke, Parkinson's disease, and others have generally been regarded as neurological condi- tions, it is important that investigators inter— ested in the pathogenesis of the psychiatric disorders vigorously pursue this line of research. Each of these conditions has psychiatric symp-— toms associated with it, and frequently these are the complaints that bring the patients to medical attention. In addition, the discrete and definable nature of the neuropathology can provide a be- ginning for investigations of the neural basis of the psychiatric symptoms seen in these condi- tions. So, for instance, while depressions asso-— ciated with stroke have not been of particular interest to neurological investigators, poststroke depression may provide an important lead for mental health researchers—-helping to identify brain regions relevant to the regulation of mood states. Pharmacological Models. The utility of animal models of pathology has been well established. Models of medical diseases with the attribute of a close correspondence between the etiology of the human disease and that of the model are termed homologous. Other models are termed isomorphic; that is, despite similarities between the form of the model and the human condition, the cause of the condition in the animal may be quite different from the cause in the human. Models used in the study of mental diseases are usually nonhomologous, because of the uncer-— tainty about the etiology of the mental disorder, and nonisomorphic, because the mental disorder in humans is characterized by clinical symptoms and verbal reports that cannot be confirmed in animals. Many of the animal models of schizophrenia are drug-related (see chapter IV). By far the greatest amount of research has focused on the amphetamine model of schizophrenia. In addi- tion, hallucinogens have been used because of their strong association with the production of hallucinations in humans. Other drugs also have been used to create a model of schizophrenia, including 6-hydroxydopamine. ''Animal models of affective disorders have been produced by both pharmacological and so-— cial manipulations. Many of the pharmacological models have been derived from the observations that the Rauwolfia alkaloid reserpine can pro- duce a depressive disorder in humans, and anti- depressants can reverse the behavioral effects induced by reserpine in rats. In addition, other drugs that inhibit biogenic amine synthesis have been used to create animal models of depression. Depletion of serotonin using para—chlorol- phenylalanine has prompted the hypothesis that hypersensitive serotonergic receptors may be responsible for the depression that occurs fol- lowing stress. Social Models. The socially induced models of depression have been of two major types, social isolation and learned helplessness. Animals sub— jected to prolonged isolation develop acute dis-— tress, agitation, and subsequent retardation of activity and weight loss; death is not an uncom-— mon outcome. The term learned helplessness has been used to describe the acquisition and per— formance deficits seen along with changes in activity, food intake, and weight following treatment with inescapable aperiodic shock. Both the isolation and the learned helplessness models and the despair model of depression have been found to improve with tricyclic antide- pressants and/or, in some cases, electrocon— vulsive shock. Animal models have been proposed for the induction of mania using 6—hydroxydopamine and for the switch into mania following microin- fusions of beta endorphin. Other models have looked at cocaine-induced psychosis and the role of hormonal influences on experimental models of neuropsychiatric disorders. Broadly defined, psychopathology can be characterized by disturbances in cognition (at-— tention, learning, and memory), affective states, sexual function, motivation, perception, and circadian or other cyclic states. Consequently, there would appear to be greater promise in studies that attempt to model clearly defined symptoms of complex human psychopathological states rather than full-blown syndromes. For example, attention deficits can be studied and modeled in animals as a core symptom of schizophrenia; changes in sexual behavior re— sulting from hormone treatments as components of sexual dysfunctions in humans; and disturb- ances in sleep and other cyclic states as symp-— toms of depression. Investigators also are studying rhythms in various patient groups. Patients with either bi- polar or unipolar depression have been studied in temporal isolation facilities and during en- trainment in hospital research environments. Studies indicate that patients with major de- pressive disorders develop a significant altera— 74 tion in phase, amplitude, and wave shape of a cluster of circadian chronophysiological func-— tions, and that these abnormalities reverse toward normal during clinical remission or treated recovery. There is strong evidence that manipulation of the sleep-wake cycle in these patients may dramatically, but transiently, ame— liorate the depressed symptoms. Recent studies have shown that total sleep deprivation for one night, deprivation of only the latter half of the nocturnal sleep episode, selective REM depri- vation, and a phase advance of the nocturnal sleep episode all produce significant but short— lived improvement of depressive symptoms. Also being studied are patients with two sleep disorders classified as the delayed sleep phase syndrome (DSPS) and the hypernychthemeral syndrome. DSPS patients sleep normally but have great difficulty falling asleep at the time they wish (a chronic delay), even though they are capable of entrainment to a 24-hour sleep-wake cycle. The hypernychthemeral syndrome is present in patients who are unable to entrain to a 24-hour schedule and develop instead a nonen— trained sleep-wake rhythm of, typically, ap-— proximately 25 hours. Patients with DSPS have been successfully treated by a form of chrono— therapy, which progressively delays their sleep time by an amount each day within their range of entrainment until the daily sleep episode is placed and held at a desired phase of clock time. Another area of basic and clinical research deserves mention because of its importance in clinical psychiatry. Recent studies in animals and humans have emphasized the important ef-— fect of certain psychoactive drugs on the period and phase of circadian rhythms. For example, lithium, clorgyline, and imipramine will lengthen the period of activity-rest cycles in rodents. The importance of elapsed waking time (or even elapsed time alone) leading to increasing sleepiness and an increased amount of stages 3 and 4 sleep is also recognized. Most recent con— ceptual models of sleep-wake cycle organization provide for that component. There are contro- versies regarding the number of coupled oscilla— tors needed in a model to fit what is known, and whether specific mathematical nonlinear coupled oscillator models are appropriate, e.g., Van—der— Pol oscillators. The use of mathematical models such as coupled Van-der-—Pol oscillators will allow investigators to do simulations using com— puter displays and then to test them in the chro- nophysiology laboratory. More generally, tests of the various mathematical single or multi- oscillator models by systematically manipulating sleep-wake timing will provide important new insights into the properties of the underlying biological CNS oscillators. Only a few studies have explored the effects of maturation and aging on the oscillators’ underlying rhythmic processes. Data on the human aging process suggest that there are changes in the period of ''sleep-wake and temperature rhythms, the amount of stages 3 and 4 sleep, and the ampli- tude and frequency of growth hormone release. A previous report that older individuals have a high probability of desynchronization during nonentrainment has not been supported by recent results. The importance of ultradian rhythms and the direct superimposed influence of behavioral ac-— tivities has also recently been recognized. In normal adults, going to sleep each biological day, whether entrained or free-running, almost al- ways results in the development of stages 3 and 4 sleep, the release of growth hormone, the in— hibition of cortisol secretion, and a drop of body temperature. These changes occur within the first 1 to 2 hours after sleep onset. It is now clear that animal models for the study of the neural substrates of behavior need not be limited to primate or even vertebrate species but can, in certain circumstances, be carried out in invertebrates as well. The models selected for study should not only be capable of generating valid extrapolations to human psy- chopathologies but should also be amenable to interdisciplinary study. That is, such animal models should be chosen so that they can be examined from morphological, electrophys-— iological, neurochemical, and molecular bio-— logical approaches. , Recommendations for the Future Explicit examples of research needs and op— portunities are replete throughout this chapter. Summarized below are those of the panels' rec-— ommendations that deal specifically with the neural basis of psychopathology. ® Development of definitive tests of the ex- isting hypotheses of the major mental ill- nesses by: — Studies of normal humans that examine perceptual, cognitive, linguistic, and at-— tentional processes utilizing such tech-— niques as activity imaging (NMR, PET) and noninvasive electrophysiological responses (ERPs, magnetoencephalography) 7S - Studies of receptor properties through the use of specific ligands and of receptor states in well-defined human autopsy material - Development of genetic markers for mapping the loci of genes contributing to psychiatric illnesses Further definition of the neuronal circuitry of transmitter systems of interest in non- human species and application of the knowledge gained to defining these circuits in autopsy material from control and men— tally ill subjects Studies of living subjects with well-defined CNS lesions who have associated psychiatric symptoms in order to understand the path- ophysiology and neuropathology of specific psychiatric symptoms Elucidation of the roles of various plastic and homeostatic systems in the brain that mediate behavior and the maintenance of mental normalcy Studies on the maternal and environmental influences on the development and organi- zation of transmitter systems of interest in mental illness Clarification of the interactions between the nervous, endocrine, and immune systems Elucidation of the underlying neural mech-— anisms of short- and long-term memory Studies on nutritional requirements and mechanisms for proper development and function of the CNS Elucidation of the neural bases of cognition incorporating the most productive new concepts in cognitive psychology Studies of prenatal development and fetal behavior, along with the development of appropriate methodology for use in this area ''ACh: AChR: ACTH: AMP: ATP: AVP: Catt; CCK: CNS: cDNA: CSF: CRF: cAMP: cGMP: DA: DNA: DOPA: DSPS: 2-DG: ECF: ECT: EDso: ABBREVIATIONS Acetylcholine Acetylcholine receptor Adrenocorticotropic hormone Adenosine monophosphate Adenosine triphosphate Vasopressin Maximum ligand binding Calcium Cholecystokinin Central nervous system Complementary deoxyribonucleic acid Cerebrospinal fluid Corticotropin releasing factor Cyclic adenosine monophosphate Cyclic guanosine monophosphate Dopamine Deoxyriboneucleic acid Dihydroxyphenylalanine Delayed sleep phase syndrome 2-—deoxyglucose Epinephrine Extracellular fluid Electroconvulsive therapy Effective dose in SO percent of subjects 76 EEG: ELH: EP: ERP: GABA: GMP: GnRH: GTP: Ht: HCO3: HD: SHEL: ISMs: LH: LHRH: Lit; LSD: MAM: MEG: MHPG: Mr: Electroencephalogram Egg-laying hormone Evoked potential Event related brain potential Gamma aminobutyric acid Guanosine monophosphate Gonadotropin releasing hormone Guanosine triphosphate Histamine Hydrogen ion Bicarbonate ion Huntington's disease 5-hydroxytryptamine (serotonin) Ion-selective microelectrodes Ligand affinity Luteinizing hormone Luteinizing hormone releasing hormone Lithium ion Lysergic acid diethylamide Methylazoxymethanol Magnetoencephalography 3-—methoxy--4—hydroxy-—phenylglycol Relative molecular weight ''mRNA: N-CAM: NDH: NE: NGF: NMR: NREM: 6-OHDA: PCP: PET: Messenger ribonucleic acid Neural-cell adhesion molecule Neurodepressing hormone Norepinephrine Nerve growth factor Nuclear magnetic resonance Nonrapid eye movement 6—hydroxydopamine Phencyclidine Positron emission tomography Yu. Ss. PKC: PNS: POMC: REM: RNA: SCP: SCN: TRH: VIP: 77 Protein kinase catalytic subunit Peripheral nervous system Pro—opiomelanocortin Rapid eye movement Ribonucleic acid Small cardioactive peptide Suprachiasmatic nucleus Thyrotropin releasing hormone Vasoactive intestinal peptide GOVERNMENT PRINTING OFFICE: 1984-454°462/19385 '' '' '' '' EVALUATION PANEL MEMBERS - continued Robert Y. Moore, M.D., Ph.D. State University of New York, Stony Brook Ronald W. Oppenheim, Ph.D. Bowman Gray School of Medicine James W. Patrick, Ph.D. Salk Institute Paul H. Patterson, Ph.D. California Institute of Technology Ernest J. Peck, Jr., Ph.D. University of Arkansas for Medical Sciences Harvey B. Pollard, M.D., Ph.D. National Institute of Arthritis, Diabetes, Digestive and Kidney Diseases Michael I. Posner, Ph.D. University of Oregon Michael A. Raftery, Ph.D., D.Sc. California Institute of Technology Marcus E. Raichle, M.D. Washington University School of Medicine Allan Rechtschaffen, Ph.D. University of Chicago School of Medicine Russel J. Reiter, Ph.D. University of Texas Health Science Center, San Antonio Elliott Richelson, M.D. Mayo Foundation Trevor Robbins, Ph.D. University of Cambridge James L. Roberts, Ph.D. Columbia University College of Physicians and Surgeons Robert G. Robinson, M.D. Johns Hopkins University School of Medicine Evelyn Satinoff, Ph.D. University of Illinois, Champaign Peter Schiller, Ph.D. Massachusetts Institute of Technology David Schubert, Ph.D. Salk Institute Ellen K. Silbergeld, Ph.D. Environmental Defense Fund Charles B. Smith, M.D., Ph.D. University of Michigan School of Medicine Gerard P. Smith, M.D. Cornell University Medical Center Larry R. Squire, Ph.D. University of California, San Diego Charles F. Stevens, M.D., Ph.D. Yale University School of Medicine Felix Strumwasser, Ph.D. Boston University John F. Tallman, Ph.D. Yale University School of Medicine Richard F. Thompson, Ph.D. Stanford University James W. Truman, Ph.D. University of Washington Fred W. Turek, Ph.D. Northwestern University Sylvio S. Varon, M.D. University of California, San Diego Stanley J. Watson, Jr., M.D., Ph.D. University of Michigan School of Medicine Norman Weiner, M.D. University of Colorado School of Medicine Elliott D. Weitzman, M.D. Cornell University Medical Center Donald J. Woodward, Ph.D. University of Texas Health Science Center, Dallas NEUROSCIENCES RESEARCH BRANCH BRANCH CHIEF PROFESSIONAL STAFF Stephen H. Koslow, Ph.D. Niles Bernick, Ph.D. Biobehavioral Research Section Ronald Schoenfeld, Ph.D. Psychopharmacology Research Section —J. Stephen Kennedy, Ph.D. Sleep and Biological Rhythms Program —Albert A. Manian, Ph.D. Chemical Synthesis Program Steven Zalcman, M.D. Neurobiological Research Section Muriel Reich Asher Research Assistant ''DHHS Publication No. (ADM) 84-1363 Printed 1984 ''