2
Advancing Neuroscience in the Decade of the Brain
Not very long ago, the feasibility of mapping the distinguishable regions of the human brain in relation to their functional roles seemed remote. With the tremendous advances in neuroscience in the past two decades, however, the opportunity now exists to approach the integrated understanding of brain structure and functioning necessary to clarify the neurobiological basis of human thought and emotion and to discern the mechanisms that underlie sensory perception and locomotor functions. Many of the intricate anatomical connections of the brain are being defined in great detail. New capabilities have emerged to identify and describe the biochemical, molecular, and genetic mechanisms that determine brain structure and functions. The activity of the human brain during mental activity can be measured and visualized. It is even becoming possible to monitor simultaneously the activity of many neurons within complex neural networks during discrete behaviors. The challenge now is to establish a comprehensive initiative that will increase the ability of neuroscientists to make discoveries about the brain and to apply this knowledge to the many mental and neurological disorders that affect humankind.
The progress made in this area has occurred primarily through the concerted efforts of increasing numbers of individual investigators, working mostly in small groups on highly specific projects. The body of information gained through such efforts has grown in a piecemeal fashion; it has now reached a point of limitation, in terms of its usefulness, because the mass of information is so great and its dissemination so poorly coordinated that critical data are often difficult to recover and
define. Indeed, this approach to neuroscience research, which was so successful in the past, may soon limit advances in the same way that a single surveyor who charts a field cannot hope to map a continent without a coordinated plan involving other mappers. A Brain Mapping Initiative could identify those aspects of information exchange infrastructure that are critical to addressing a broader goal, one that will include the advantages of single-investigator projects and yet also yield the benefits of a larger, coordinated program. The Brain Mapping Initiative is intended to subsume all the proposed aspects of a National Neural Circuitry Database outlined in the charge to this committee. It is also designed to express explicitly the goals of the proposed effort and reflect more adequately the complex of electronic and digital resources that will be required.
A consensus is emerging that the initial steps can now be taken toward the global task of understanding brain structure and functioning. The impact of digital computer technology began in the physical sciences three or four decades ago and led to such current large-scale efforts as the supercollider, space telescope, and interplanetary probes. In neuroscience, the increasing availability of new enabling technologies is likely to have similar, far-reaching impacts. The development of high-density memory chips and the latest generation of microprocessors provides a key stimulus to accelerated development of image analysis graphics and image manipulation—a set of capabilities known as visualization computing (McCormick et al., 1987). The emergence of parallel processing, scientific visualization workstations, and high-capacity digital communications may provide the technical support needed to conduct coordinated projects in neuroscience.
A comprehensive, coordinated effort to understand basic organizational patterns of brain connections needs to be undertaken. This effort should include a definition of the chemical identity of neuronal populations and a description of neuronal structure and neuronal circuit organization in each region in sufficient detail to clarify the computational processes involved. The pace of future advances in neuroscience will depend on critical choices, which need to be made now, regarding the handling of information to be gathered in the future. At issue is whether neuroscientists will embark on a large-scale effort to develop and integrate new forms of technology for acquiring and managing information.
Complexity and the Need for Information Management
The brain exhibits by far the greatest complexity of any of the organs of the human body. Indeed, there is reason to believe that a sub-
stantial percentage of the human genome plays a role in creating neural structure, connectional patterns, and neurochemistry. The brain's responsibilities are numerous and include initiation of movement, mediation of sensation, and regulation of other body organs. Its functioning provides the basis of perception and cognition and underlies emotion and creative expression. Although described as a master computer, the brain is fundamentally more complex and its processes far more subtle than those of any current computer design. With the advantage of parallel operation of neuronal populations, the brain manages and controls a wide variety of tasks simultaneously, reliably, and with rapid precision. Indeed, much of the brain's work proceeds even in the absence of an individual's conscious awareness.
All brain activity results from electrical and chemical communication among neurons (the primary signaling cells of the brain), each of which can communicate with other neurons using signals at rates of up to 1,000 events (impulses) per second. To understand the brain, neuroscientists must measure and analyze the rapid changes in neuronal signaling activity that occur over the vast networks of cells and connections. The scope of this endeavor is immense. It is estimated that the human brain contains more than 100 billion neurons, and each neuron maintains an average of about 1,000 connections, called synapses, with other neurons. Some neurons have as many as 200,000 synapses. During each moment of daily life, neural signals may be transmitted across any of approximately 100 trillion synapses.
Improvements in human health require greater integration and synthesis of knowledge than can be gained simply by describing the intricate, extensive circuitry that characterizes a normal or average brain. To treat schizophrenia, Alzheimer's disease, brain injury, degenerative diseases, developmental deficits, and chemical dependency, it will be necessary to understand the biochemical reactions and specialized neurochemistry expressed by different types of neurons. In addition, neuroscience must take into account the differences among individuals that arise from environmental as well as hereditary factors. Such differences range from subtle genetic variations in small populations of neurons to differing developmental states and wide-ranging perturbations of major brain systems.
In many ways, because of the brain's organization, neuroscience research is obliged to proceed along a multilevel hierarchy, from behavior to molecular interactions(see Chapter 3 ). Each level of research uses numerous techniques that are specifically designed to collect information appropriate to that level. But information from one level often has important implications for knowledge at other levels. Effective synthesis of the data depends on the traditional academic skills of
the scholar; it is becoming increasingly difficult, however, for neuroscientists to identify, manage, and process all the information needed for the efficient design and conduct of their research. Even with well-developed traditional academic skills, the neuroscientist clearly needs new information organization and access technologies. Electronic information management provides assistance at a high level of synthesis—in much the same way that hand calculators and word processors facilitate tasks like balancing a budget or creating a document.
To draw an analogy, neuroscientists are like puzzle builders who must integrate and fit together numerous small pieces of information derived from the hundreds of available techniques. Metaphorically, the challenge is to assemble a multimillion-piece, three-dimensional puzzle, starting with an incomplete set of pieces (many key facts about the brain remain unknown) and incomplete knowledge of how different pieces relate to one another (seemingly disparate facts are often closely related). In addition, the puzzle that represents the brain must depict more than simply a structure. It must integrate structure with function, function with chemistry, and chemistry with genetic mechanisms. Generating the data that represent the individual pieces of the puzzle is difficult in and of itself: the techniques available to neuroscientists all have inherent limitations and usually produce only incremental bits of information, although every year more bits are added and new puzzle pieces identified. The next decade is expected to provide enough new pieces to allow for the discovery of exciting new ways to understand, protect, and restore brain functioning. But innovative methods of information assembly and integration will be as important as the discovery of the individual pieces.
In science, the conventional way to gain access to relevant information generated outside one's own laboratory is through the study of scientific literature and by formal and informal communication with other scientists. As the number of publications and literature sources grows, the use of such resources becomes more and more expensive, inefficient, and incomplete. It now takes an inordinate amount of time to sift through the various conventional publications. The process usually requires a narrowly focused search strategy, which yields less information than is actually available. Eventually, the process is self limiting. To pursue the puzzle analogy, it is like trying to assemble the three-dimensional puzzle by using two-dimensional sketches of the pieces, rather than being able to hold the original three-dimensional pieces. Furthermore, because the pieces are scattered in various locations, most of one 's time is spent in looking for the individual pieces and not in trying to understand how they are interrelated.
For many years, a number of neuroscientists have envisioned alternative strategies for coping with the problem of information management, which is distinct from that of acquiring more data. Information management is aimed toward realizing the greatest benefit from the data that already exist. Two major kinds of information management are the most critical for neuroscience: (1) databases, which allow at least partial automation of the task of relating diverse data types in a systematic, efficient manner, and (2) the capacity for visualization of structures in three dimensions. Visualization computing, whether by machine or by the scholar's brain, captures the discoverable architecture of the brain and relates this architecture to brain functioning.
Scientists mapping the human genome or sequencing biologically important proteins have been using databases for years (Smith, 1990; Vela, 1990). These databases contain precise information, obtained directly from experimental investigation, regarding the sequence of base pairs of specific genes, their locations on specific chromosomes, and the amino acid sequences of known and newly discovered proteins. Investigators consider these databases invaluable for a host of reasons. Most commonly, researchers check a newly charted protein sequence against thousands of other known sequences to see if their new sequence is related to other known proteins. Similarly, the base pair sequences of genes of one species are routinely checked against sequences from other species to determine the level of homology of specific genes across species. Beyond simple comparison, knowledge of base pairs and protein sequences can provide key information regarding biological functions. Thus, protein and gene mapping databases are more than simply references; they are quickly becoming essential components of the scientific process of discovery (see Box 2-1 ; DeLisi, 1988; Colwell, 1989).
The experience of the protein and gene mapping community is an instructive example of the benefits to be gained from incorporating database and information technology into biomedical research efforts. There are major differences, however, that must be taken into account when considering how to extend this approach to the field of neuroscience. Chief among these differences is that of complexity. Although proteins exhibit tremendous diversity, much of this diversity is represented by the linear sequence of amino acids that make up each protein. Such linear arrays are easy to store in traditional databases and also represent the key information for genes. All proteins also have a three-dimensional structure, however. Computerized molecular modeling has been very successful in displaying molecular structures in three dimensions (Howard Hughes Medical Institute, 1990), but it is not yet possible to incorporate these three-dimensional models into
BOX 2-1 THE GENE FOR NEUROFIBROMATOSIS 1 First described in the late nineteenth century, neurofibromatosis 1 (NF1) is an inherited condition affecting 1 in every 3,500 people. The two most common symptoms of NF1 are areas of dark pigmentation on the skin, and benign but often disfiguring tumors, called neurofibromas, beneath the skin. A wide range of other, less common symptoms appear in connection with NF1, and these include learning disabilities, mental retardation, malignancies of the nervous system, scoliosis, and other bone abnormalities. In 1987, James Gusella and his colleagues at Massachusetts General Hospital (Seizinger et al., 1987) and Ray White and his research team at the University of Utah (Barker et al., 1987), localized the gene responsible for causing NF1 to chromosome 17. The abnormal gene was subsequently isolated and sequenced by White's team and by Francis Collins and his coworkers at the University of Michigan (Lewis, 1990). The next step was to translate the genetic code into an amino acid sequence, which was then fed into a computer and matched against databases of other known sequences. Similarities were found between the protein sequence of NF1 and those of two other genes, one being a gene that codes for a group of proteins called GAP proteins, which are thought to suppress tumor growth, and the other being a tumor-suppressing gene in yeast. The homologies seem logical—altered GAP protein function could certainly have manifestations similar to those observed in NF1. Although many important details remain to be filled in, what scientists already know about GAP proteins and yeast genes will provide a springboard for researchers striving to understand neurofibromatosis. The knowledge base established by White, Collins, Gusella, and their colleagues is likely to attract researchers from many disciplines that have not traditionally studied neurofibromatosis—cancer researchers in particular, because NF1 now joins the growing list of potential tumor-suppressor genes. A complete solution to the NF1 puzzle is unlikely in the very near future, but discovery of the NF1 gene and its protein product promises to speed development of prenatal diagnosis of and possibly treatment for NF1. |
traditional databases. In contrast to molecular biology, understanding the diversity inherent in brain structure and functioning requires the three-dimensional display of many types of experimental information. Because the fundamental complexity of neuroscience data exceeds that of molecular biology data, the development of information management technologies to enhance neuroscience research will present great challenges. Yet such challenges are approachable, and the potential benefits to greater understanding of the human brain are compelling reasons for confronting them.
Examples of the Value of Integrating Knowledge to Solve Problems
Achieving advances in neuroscience is a difficult task, particularly those advances that yield improvements in the treatment and understanding of neurological, developmental, and mental diseases. Integration of knowledge across the areas of investigation at each level of the neural hierarchy is required. The importance of applying these diverse methodologies to discern how the brain functions can be illustrated by the following explicit examples of the hierarchical organization within neuroscience (see also Chapter 3 ).
Vision: How we see affects how we think
The visual system, one of the most intensively studied systems of the brain, reveals remarkable complexity in organization and functioning. Because it occurs so effortlessly, vision is something we often take for granted and consider to be a fairly simple matter. In fact, however, the brain conducts extensive processing of all visual information to allow for pattern recognition, interpretation, and appreciation of such concepts as beauty. The neural underpinnings of these abilities are contained within a rich, intricate network of pathways that involve billions of individual neurons. The system also includes muscles and motor regions of the brain to point the eyes toward areas of interest and to track objects in motion (see Box 2-2 ).
Investigations of the visual system have used a wide range of techniques that have contributed substantial knowledge regarding the anatomy, biochemistry, and physiology of the system. The information that reaches the eyes is first processed by a thin layer of cells that make up the retina at the back of the eye. Through exquisitely sensitive signal transduction mechanisms, retinal cells respond to a range of inputs, including color. Coordination among the cells in the retina also provides the first filtering of incoming information by the balancing of excitatory and inhibitory processes. From the retina, information is transmitted to a pair of nuclei on each side of the brain, called the lateral geniculate nuclei, in a highly ordered pattern so that each half of the visual field is precisely mapped onto the lateral geniculate on the opposite side of the brain. So well understood is this map that differing kinds of visual deficits can indicate the exact location of damage (a frequent occurrence following stroke or as a result of the pressure of a growing tumor) along the pathways of the visual system. The lateral geniculate nuclei relay visual information to the primary visual cortex. Again, the information is mapped in a highly ordered pattern with segregation of the inputs from each eye.
BOX 2-2 SOMETIMES THE BRAIN LEARNS TO IGNORE VISUAL INPUT To focus on objects in the visual field, the muscles of both eyes are normally coordinated so that both eyes move in unison. But abnormal eye muscle development in one eye can cause misalignment of the two eyes, often resulting in a condition known as strabismus. A child born with this condition will favor one eye to avoid double vision, and, over time, visual acuity in the less favored eye deteriorates. Left uncorrected, the child becomes blind in that eye. Before the neural basis of binocular vision was understood, procedures to correct strabismus were typically done at age 8 or 9, long after permanent deterioration had occurred. Today, the condition is corrected very early to protect such children's normal vision. |
Measurement of how neurons in the visual cortex respond to different stimuli has revealed much about the microprocessing of information. Most important, many visual neurons are selective for specific features in the visual world. For example, Nobel laureates David Hubel and Torsten Wiesel discovered that many neurons are selective for the orientation of a bar, which allows them to signal the presence of particular edges and contours (Hubel and Wiesel, 1979). Since their discovery, other researchers have described types of neurons that are sensitive to color, directional movements, and, at higher levels, even complex shapes like faces.
From the primary visual cortex, information is transmitted to many other cortical areas involved in higher aspects of visual processing. This dispersion of information is necessary for all the activities associated with vision, including reading, writing, and recognition of objects in the environment. The monkey has 32 regions of the cortex that are known to be involved in some type of visual processing ( Plate 2-1 ). Each of these regions subserves a specific function; yet they are interlinked by more than 300 separate pathways. Much visual information ultimately reaches neural centers involved in cognitive functions and emotions. Understanding these parallel, distributed pathways has important implications for understanding the deficits from brain injury, including stroke. They also help explain some of the most puzzling aspects of these problems. (For example, some stroke patients can recognize words and objects such as tables and chairs, but cannot recognize a loved one's face [Sacks, 1970]. Because each of these different stimuli present distinct features, damage to one area of the cortex may affect the perception of one constellation of features but not others.)
Cognition and recognition are not the only results of visual sensation. The retina also projects visual information to nuclei that are involved not in feature analysis but in controlling the eye muscles. These pathways mediate the turning of the eyes in the direction of a stimulus; the coordinated, simultaneous movement of both eyes; and the ability to track moving objects in the visual field. Again, certain kinds of damage affect these pathways and may result in debilitating deficits.
As in all neural systems, the functioning of the visual system relies on the coordinated activity of neurons that communicate with each other using electrical and chemical signals generated by hundreds of different molecules, including neurotransmitters, second messengers, and signal transduction molecules. Many of these molecules have been found in distinctive patterns in specific areas of the visual system. The combination of complex physiological responses, complex anatomical pathways, and hundreds of neurochemical interactions creates a system that is extremely difficult to disentangle. Despite all the information we possess, we still do not understand the fundamental nature of visual perception, nor do we understand the specific computations carried out in the vast networks at each stage of information processing.
Finding answers to these challenging questions will require not only additional experimental data but new strategies for assembling the available information to facilitate integration of the diverse data types. To accomplish this integration, the use of three-dimensional graphics, sophisticated database technology, and greater levels of electronic communication among investigators will be critical. For example, maps of the anatomical pathways must be overlaid with neurochemical maps to understand how the wide repertoire of chemical modulatory influences may be related to different aspects of visual perception and visual-motor coordination. Computer simulations and realistic computational models, based on physiological and anatomical data, are needed to understand the specific signal-processing strategies used in various regions of the visual system. Such computational studies are also of interest to engineers and computer scientists in the construction of new types of sensors and signal detectors for nonbiological uses.
It is clear that the full potential of visual system research has yet to be realized. Progress in this field will contribute to the ability to treat or compensate for visual deficits caused by brain injury. In addition, this work can contribute to the amelioration of blindness and diseases of the visual system, including glaucoma, diabetic retinopathy, and inherited retinal degenerations. The mapping and physiological characterization of the multiple visual system path-
ways also have direct relevance for understanding other portions of the brain, especially other portions of the cerebral cortex.
Substance abuse: The search for the biology of self-destruction
A large fraction of the U.S. population uses substances that are injurious to personal health. These substances range from legally approved drugs (e.g., nicotine and alcohol) to illegal drugs (e.g., cocaine and heroin) that are accompanied by grave societal costs such as violent crime and infant mortality. The key to understanding and dealing with substance abuse lies in the brain. One path of research has already led to the critically important discovery that there are cell-surface receptors not only for neurotransmitters but also for many drugs, including nicotine, marijuana, heroin and other opiate drugs, and benzodiazepines, such as the tranquilizer Valium® (Plate 2-2). Often, certain drugs bind to the same receptor to which a particular neurotransmitter binds. For example, the two kinds of acetylcholine receptors were first distinguished by the fact that one bound nicotine and the other did not. Such receptors have been mapped to specific nuclei of the brain. Thus, the notion that certain drugs have specific actions in discrete brain regions offers a possible route for disrupting the effects of those drugs.
But the problem goes much deeper. For example, alcohol does not act at specific receptor sites; its effects are much more general throughout the brain. Furthermore, all psychoactive substances produce experiences that provide some type of motivation or pleasure that can lead to addiction. Therefore, addiction to medically harmful substances is directly related to the involvement of brain systems that mediate the normal pleasurable experiences associated with many day-to-day activities such as eating, drinking, and socialization. Neuroscientists do not completely understand the location, organization, and chemistry of the system in the brain that underlies positive emotions and pleasure and, with that, the pleasurable, motivating effects of drugs and alcohol. Certain fascinating clues, however, have been found. For example, animals will self-administer cocaine so avidly that they totally ignore food; yet this behavior can be completely reversed if the neurotransmitter dopamine is removed from an area of the brain known as the nucleus accumbens. No conceptual framework similar to those that guide our understanding of the visual and motor systems has emerged to explain this phenomenon. Many scientists hypothesize that a number of brain regions in the limbic system and hypothalamus are involved, but the connections between these regions are extremely complex and not well understood.
Two types of databases could greatly enhance research into the biology of substance abuse. Databases that allow for the easy comparison of brain distribution patterns of receptors for the many substances that are abused would be invaluable. Their usefulness would be particularly great if distribution patterns could be compared with the known circuitry of the brain, with the associated distribution patterns of known neurotransmitters, receptors, and ion channels, and with the results of behavioral studies related to the neural mechanisms of addiction. No conventional library system currently allows for this level of analysis of visual material. A second, related kind of database would contain information about the known neural pathways implicated in addiction. Such a resource could aid the development of a viable picture of the brain system involved in mediating the reinforcing aspects of drug use and in the formulation of efficient strategies for research into areas where more information is needed. Substance abuse offers a compelling example of how the availability of comprehensive databases that incorporate and integrate information from many fields, including chemical neuroanatomy, pharmacology, and biopsychology, would almost certainly speed the development of effective treatments for a widespread, serious threat to human health.
Pain: Sometimes a warning, sometimes a curse
Pain is a ubiquitous reality of life. We need pain to recognize sprained ankles, overstressed back muscles, kidney stones, infections, and a host of other ills. This useful warning system can go awry, however, and become an intractable barrier to a happy, productive life (Plate 2-3 ). The neural underpinnings of pain involve almost every region of the brain, spinal cord, and peripheral nervous system. Moreover, these neural mechanisms interact with other systems of the body, including the immune system and certain endocrine glands, and are even affected by changes in the metabolic status of the individual (e.g., as in diabetes). Pain is thus a remarkably complex process about which much remains to be learned.
In its simplest form, pain results from the activation of peripheral nerves by certain types of stimuli impinging on the skin and internal organs. This neural information is transmitted to and processed by specific parts of the spinal cord. From the spinal cord, information about injury is transmitted to numerous brain regions, including specific relay nuclei that, in turn, transmit it to the sensory part of the cerebral cortex and to nonspecific nuclei that disperse the information through their abundant connections to other brain regions. In-
formation regarding a painful stimulus reaches brain regions that are involved in emotion, sensory perception, body movement, and hormonal release. In addition, the information reaches brain regions that connect back to the spinal cord and that are capable of modulating the activity of spinal neurons. These complex connections form the anatomical basis for systems that can activate or inhibit pain. The pain system also utilizes scores of different neurotransmitters and an interesting group of peptides that include endorphin and enkephalin, which are similar to morphine.
The intricacy of these connections explains in part some of the most fascinating and worrisome aspects of pain perception. For example, it is common for wounded people to walk considerable distances to reach help without conscious awareness of their pain and for people with malignant cancerous growths to be free of pain in the cancer 's early stages (possibly because of the activation of pain-inhibiting systems). On the other hand, people who have had amputations, including mastectomy, often feel pain in the amputated body part, despite its absence (possibly because of pain-activating systems). The intricacy of the pain system also accounts for the wide variety of kinds of pain people experience, of kinds of treatments that are effective, and of scientific and clinical disciplines that deal with pain.
The need for enhanced communication in neuroscientific disciplines is well illustrated by the example of pain research and treatment. In the clinical arena, neurosurgeons have had a long-standing interest in finding ways to disrupt appropriate brain areas to alleviate intractable pain. The electrical stimulation or selective destruction of such areas has met with limited success, sometimes utter failure; in the worst-case scenario, the pain returns and is greater than before the surgery. Further advances in these and other techniques depend on expansion of knowledge from anatomical, electrophysiological, and neurochemical studies in animals.
The experience and knowledge of other disciplines are also important. For example, anesthesiologists deal with pain on an everyday basis, but the responsibility for pain management typically changes from the anesthesiologist to other health care providers as a patient is moved from the recovery or delivery room to a regular medical/ surgical ward. As a result, widely divergent strategies are used. Increased communication among these various care providers could lead to more effective pain management and the ability to use pain (and the lack of it) as a diagnostic tool. As it becomes clearer that pain management has a substantial impact on recovery following surgery, communication between health care providers and basic researchers should be increased to promote better information transfer.
Basic pain research has generated an impressive amount of knowledge about the brain and the way it deals with pain. Recent work has begun to characterize the connectivity patterns of spinal and brainstem pathways that mediate pain, as well as their biochemical and physiological qualities. It is becoming increasingly evident that the neural mechanisms underlying musculoskeletal, cutaneous (skin), and visceral (internal organs) pain are different, and these differences are under active investigation. Differences in chronic and acute pain also are being intensely studied. Investigators are defining as well the actions of various immune system products at the site of injury and their contribution to pain. A very recent development has been the use of animal models for pain caused by arthritis and peripheral nerve injury. This work has brought the first hints that damage to the nerves can result in permanent changes in the spinal cord, and possibly in other brain regions, that may well be responsible for intractable pain syndromes. Prevention of these changes by early intervention may help to alleviate a sizable proportion of the human suffering that results from this kind of pain.
Given the prevalence of pain and the societal burdens it imposes, there is a pressing need to find better ways to transfer into clinical practice what is being learned from basic research and to apply clinical observations to the design of basic research experiments. A family of computerized resources could promote these information transfers by making anatomical, biochemical, and physiological data available in a manner that integrates the information graphically, cross-references data from other disciplines, and provides relevant clinical observations. A repository for various pain diagnostic and treatment strategies and observed outcomes would also be useful. Finding answers to the many unresolved questions about pain and its varied pathologies will depend, in part, on several communication-related factors including the most efficient use of the information that exists now and the widest dissemination of the information that will be generated in the future.
Schizophrenia: Broken minds, shattered dreams
For the 2 million Americans who suffer from schizophrenia, the world is often terrifying, confusing, and bitterly lonely. This devastating disease of the brain most often strikes people in adolescence and young adulthood and is expressed in a variety of forms, depending on the specific constellation of symptoms that occur. These forms range from those with predominantly “negative” (loss of function) symptoms to those with mainly “positive” symptoms (exaggerated
or distorted brain functions). Thus, a person with schizophrenia will exhibit some combination of the following: hallucinations and delusions, blunted or inappropriate emotional expression, inability to derive pleasure from normal experiences, cognitive difficulties, and abnormal socialization. Major efforts are under way in neuroscience to discover the causes of schizophrenia, to understand the mechanisms of the disease, and to find better treatments for those whose lives are burdened by this illness.
Researchers are pursuing three major areas of investigation: neurochemical abnormalities, structural and functional brain abnormalities, and possible genetic and environmental causes of the disease. The discovery that certain drugs can alleviate the hallucinations and delusions of schizophrenics provided strong evidence that a neurochemical imbalance was important in the disease. Because all effective drugs were found to decrease the effects of the neurotransmitter dopamine, research is now focusing on the mechanisms of dopamine action in the brain and on mapping the areas of the brain that use dopamine in neural transmission. Unlike Parkinson's disease, a condition in which a specific group of dopamine-containing neurons die, people with schizophrenia have no single, well-defined area of the brain that is damaged; rather, anatomical studies suggest that many dopamine-containing brain areas are affected. In addition, unlike Parkinson's disease, which exhibits a loss of dopamine owing to neuronal death, schizophrenia displays an apparent excess of dopamine activity. This excess activity may be due to an abnormally high number of a specific type of cell-surface receptor for dopamine, the D2 receptor. (Receptor subtypes are themselves a subject of intense interest because each subtype behaves differently; thus, their existence adds ever deeper levels of complexity to the questions regarding the precise role of dopamine in schizophrenia.)
Further research is necessary to understand how abnormalities in the dopamine system of the brain contribute to the symptoms of schizophrenia. Evidence is mounting that other neurotransmitters are also involved. For example, it is well known that dopamine blocking drugs are effective only against the positive symptoms of schizophrenia, which suggests that the blunting of affect and other negative symptoms do not result from dopamine abnormalities. Recent studies have shown that these negative symptoms can be lessened by drugs that act on another neurotransmitter, gamma aminobutyric acid. Given the variable symptomatology of schizophrenia and the complexity of the brain systems involved, it is not surprising that continued progress is accompanied by new questions regarding the role of dopamine in this disease. Nevertheless, investigations into the neu-
![](/openbook/0309044979/xhtml/images/p200046d1g39001.jpg)
FIGURE 2-1 Magnetic resonance images of the brains of two 44-year-old male identical twins, one with schizophrenia (right) and the other without (left). Arrows point to the cerebral ventricles, which are enlarged in the affected twin. Images provided by the National Institute of Mental Health.
rochemical perturbations of schizophrenia have far-reaching implications for improved treatment of the disease. Digital maps of neurochemical systems, including the locations of different receptor subtypes, and the known functions of those systems in normal and disease states would be of particular value in the study of schizophrenia.
The ability to “image” the structure and functioning of the brain is a fundamental advance that is being vigorously applied to the study of schizophrenia. Magnetic resonance imaging (MRI) has revealed some subtle and intriguing structural abnormalities. These include in some individuals abnormally large cerebral ventricles, specific spaces in the brain containing cerebrospinal fluid, and an apparent thinning of certain areas of the cortex that are involved in emotional expression (Figure 2-1 ). Nevertheless, the functional implications of these structural abnormalities have yet to be determined. Brain imaging techniques, such as positron emission tomography (PET scanning) and its variations, can be used to measure such functions as energy metabolism, blood flow, and receptor binding; localized electrical activity can be recorded using electroencephalography and magnetoencephalography. These imaging studies can be done while subjects are actively involved in specific tasks or at rest. Such studies can also shed light on brain functioning during periods of severe dysfunction or relative
remission, and during treatment with antischizophrenic medications. Studies of regional cerebral blood flow suggest decreased activation of the frontal cortex in response to certain tasks in some people with schizophrenia. PET studies of cerebral metabolism have confirmed this dampened activity of the frontal cortex and have further suggested increased activation of other brain regions. PET studies of dopamine receptor binding in schizophrenic patients are just beginning to yield interesting, although sometimes conflicting, information.
The potential of functional imaging studies to increase our knowledge of schizophrenia is great, but these studies are complicated by several factors. For example, the variable clusters of symptoms exhibited by different patients and their differing treatment histories must be carefully correlated with the specific findings from imaging experiments. In pharmacological studies, such as those that examine dopamine receptor binding, the differing properties of the drugs that are used must be well understood to interpret the results of each experiment. Databases of imaging data combined with careful annotations regarding patient history, pharmacological properties, and other variables could greatly facilitate more rapid and more meaningful conclusions from these types of studies.
Genetic studies, involving both population and molecular genetics, are integral to research on schizophrenia. Studies of families and particularly of twins clearly suggest genetic factors in the disease. Yet attempts to locate the “schizophrenia gene” have not been successful. Candidates include genes coding for the dopamine (D2) receptor or for certain enzymes involved in dopamine synthesis or degradation. The lack of success in finding a likely single gene, combined with the complicated manifestations of the disease and the different responses to treatment, suggests that schizophrenia might actually be a group of related diseases. Indeed, strong evidence supports the notion that there is probably more than one gene related to schizophrenia and that instead of being a direct cause, it confers increased susceptibility for the disease.
This increased susceptibility raises the issue of environmental factors that might precipitate expression of such a genetic predisposition. Proposed factors range from viruses, or other infectious agents, to nutritional deficits to traumatic early childhood experiences. Information about these environmental factors and the varied clinical manifestations of the disease is exceedingly difficult to correlate. Therefore, another use for sophisticated information management tools is to begin to build a knowledge base about all that is known regarding the environmental history of each patient, in context with information about the pattern and prevalence of schizophrenia in
their families. In addition, once candidate genes for increased susceptibility are identified, they can be compared with the genetic maps in existing databases (see Chapter 4 ).
Schizophrenia stands as yet another example of the brain's complexity. Better answers about the causes of and effective treatments for schizophrenia are anxiously awaited by those who suffer from the disease and by their families. Although more data are needed to provide these answers, there is also a pressing need to correlate and combine more adequately the data that already exist.
The Growth of Neuroscience
Neuroscience research has grown in response to critical problems
The costs of caring for those who suffer from neurological disorders, drug abuse, and mental illness, combined with the cost of lost wages and other indirect losses from these conditions, are extremely difficult to estimate because reporting mechanisms cannot account for those who have more than one disorder (e.g., hearing loss and Parkinson 's disease) and the severity of these disorders vary widely. Yet even without such estimates we know that these disorders constitute a sizable liability to the health and well-being of U.S. society ( Table 2-1 ). Nearly 23 million Americans suffer from neurological and communicative disorders resulting from head and spinal cord injury, hearing and speech impairments, and infectious diseases of the nervous system, including acquired immune deficiency syndrome (AIDS). More than 3.5 million people suffer the effects of debilitating disorders such as Alzheimer's, Parkinson's and Huntington's diseases, and demyelinating and atrophic disorders such as multiple sclerosis and amyotrophic lateral sclerosis (National Advisory Neurological and Communicative Disorders and Stroke Council, 1989; National Institute of Neurological Disorders and Stroke, 1989). Impairment arising from alcohol abuse, drug abuse, and mental illness also takes a significant toll. It is estimated that more than 60 million Americans suffer from such mental illnesses as schizophrenia, affective disorders, anxiety disorders, various types of dementia, eating disorders, childhood and adolescent disorders, and sleep disorders, and more than 20 million Americans suffer from alcohol or drug abuse (Alcohol, Drug Abuse, and Mental Health Administration, 1990; Gerstein and Harwood, 1990; Rice et al., 1990).
Researchers have estimated the lifetime prevalence for any alcohol, drug, or mental health (ADM) disorder to be 32.7 percent (Regier et al., 1990). Moreover, there is significant comorbidity among ADM
TABLE 2-1 Prevalence of Selected Neurological Disorders, Mental Illnesses, and Alcohol and Drug Abuse
Condition |
Estimated Cases (in millions) a |
Alzheimer's disease and related dementias |
3.0 |
Brain tumors |
0.062 |
Epilepsy |
2.0 |
Multiple sclerosis |
0.131 |
Neuro-AIDS |
0.02 b |
Stroke |
1.9 |
Trauma (head and spinal cord injury) |
1.0 |
Mental illnesses |
|
Affective disorders |
27.7 |
Antisocial personality |
8.8 |
Anxiety disorders |
25.3 |
Schizophrenic disorders |
2.6 |
Alcohol and drug abuse |
|
Alcohol |
17.7 |
Drug |
4.6 |
Total |
94.8 |
a Some persons have more than one disorder concurrently (e.g., epilepsy and affective disorder), and these figures may reflect that comorbidity. b Incidence (number of new cases reported each year). SOURCE: For neurological disorders, NINDS, November 1989 and NANCDSC,January 1989; for mental illnesses, Rice et al., 1990; for alcoholand drug abuse, ADAMHA, 1990 and Gerstein and Harwood, 1990. |
disorders. For example, the Regier study also showed that 37 percent of those with an alcohol disorder had a comorbid mental disorder; of patients with a drug disorder (other than alcohol), more than half (53 percent) had a mental disorder as well. The frontline efforts now occurring to understand and treat this diverse array of diseases are headed by basic and clinical neuroscientists.
Neuroscience has grown in response to new technologies and an expanded understanding of biology
Much of the recent growth in neuroscience has been spurred by the development of important enabling technologies, which have opened entirely new areas of research. For example, the development of the
oscilloscope permitted the first accurate visualization of the electrical activity of neurons. The development of the electron microscope allowed investigators to study components of neurons too small to be seen with conventional microscopes. The development of anti-body techniques to tag specific molecules permitted the visualization and localization of specific neurochemicals. In addition, the combination of tissue culture techniques with antibody labeling and molecular biological techniques has provided critical new knowledge regarding the mechanisms of neural functioning. Finally, the development of computer technologies has played a fundamental role over the past two decades in enabling data collection and analysis of single-neuron physiological data. More recently, the development of graphic visualization techniques has permitted researchers to view the active functioning of a human brain.
Many other factors in addition to technology have contributed to the growth of neuroscience—for example, increased appreciation of the importance of the fundamental biological processes that underlie brain disorders, especially mental illness. In the past decade, the groundwork has been laid for tackling some of the most intractable neurological problems, including spinal cord injury, epilepsy, stroke, and neurodegenerative diseases. These and other achievements have led to a feast of exciting research opportunities that have attracted great numbers of new investigators to neuroscience in the past 10 years. The Society for Neuroscience membership roster currently stands at more than 17,500, a remarkable figure, given that there were only 500 charter members in 1969 and 6,350 members in 1979. (At the society 's annual meeting in 1990, almost 8,000 abstracts of individual research projects were presented.) Beyond U.S. borders, growth in international organizations is also apparent: the International Brain Research Organization boasts more than 20,000 members from 73 countries. An increase in scientific journals has accompanied this growth, with more than 200 scientific journals now exclusively devoted to neuroscience disciplines in the areas of basic science or clinical investigation.
Although growth in the field of neuroscience has contributed greatly to an understanding of the brain, it is not without its problems. The volume of knowledge and information generated from this work has been growing linearly with public and private investments in resources and with individual commitment. This expanding knowledge offers extraordinary opportunities for neuroscience to make rapid progress in the treatment and prevention of mental and neurological illnesses during the next decade. Without innovative strategies for information management, communication, and processing, however, the amount
and complexity of the data present daunting impediments to future advances.
Neuroscience research is a national priority
A comprehensive resource for managing information on brain structure and functioning is necessary to protect an already substantial public investment. One measure of the current national investment in neuroscience research is federal and nonfederal expenditures. Table 2-2a and Table 2-2b summarize the neuroscience-related expenditures of two agencies, the National Institutes of Health (NIH) and the Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA). Many other agencies in the federal biomedical research complex also support basic, as well as clinical, neuroscience research. The National Science Foundation has a long history of funding basic neuroscience research, and other federal research support derives from such agencies as the Department of Energy, the Department of Veterans Affairs, and the Office of Naval Research (Table 2-3 ). Outside the federal government, a number of private agencies and foundations, such as the Howard Hughes Medical Institute and the MacArthur Foundation, provide support for neuroscience research ( Table 2-3 ). Although it is impossible to determine the precise amount spent on neuroscience research (because of the diversity of funding sources and the inclusion of neuroscience-targeted funds in budgets that reflect support in more than one field of biomedical research), it is estimated that the total national investment is approximately $1.5 billion per year (Table 2-3 ).
When considered within the context of the massive disease burden from mental and neurological diseases, even greater investment in neuroscience research could be justified. Strategies such as the Brain Mapping Initiative are designed in part to protect current investments by maximizing the integration of information gained from neuroscience research. The long-term benefits of this research are numerous and will substantially reduce the national and international burden—both in terms of costs and human suffering—imposed by neurological and mental disorders. To hasten this progress, strategies must be developed to integrate new discoveries with the extensive body of accumulated knowledge. Such integration is needed now because the advances made by neuroscientists, especially in the past decade, have brought this research enterprise to the threshold of major breakthroughs in a number of areas, including neurogenetics, biobehavioral sciences, and neural injury. Neuroscience research is a complex field that comprises many subdisciplines, most of which share a common need for a comprehensive map of the brain in digital
TABLE 2-2a Investment in Neuroscience Research (in thousands of dollars) by the National Institutes of Health
Institute |
1988 (Actual) |
1989 (Actual) |
1990 (Estimated) |
Division of Research Resources |
30,301 |
34,365 |
28,728 |
National Cancer Institute |
34,068 |
41,378 |
42,753 |
National Eye Institute |
79,188 |
81,244 |
81,753 |
National Heart, Lung, and Blood Institute |
52,970 |
56,280 |
57,700 |
National Institute of Allergy and Infectious Diseases |
24,571 |
40,197 |
41,805 |
National Institute of Arthritis and Musculoskeletal and Skin Diseases |
628 |
1,137 |
1,255 |
National Institute of Child Health and Human Development |
78,173 |
91,032 |
95,000 |
National Institute of Dental Research |
9,337 |
8,511 |
8,755 |
National Institute of Diabetes and Digestive and Kidney Diseases |
25,500 |
26,900 |
28,100 |
National Institute of Environmental Health Sciences |
15,588 |
19,120 |
19,866 |
National Institute of General Medical Sciences |
11,400 |
11,500 |
11,750 |
National Institute of Neurological Disorders and Stroke |
458,792 |
471,632 |
490,409 |
National Institute on Aging |
63,221 |
78,573 |
84,380 |
Total |
883,707 |
961,869 |
992,254 |
SOURCE: Division of Financial Management, National Institutes ofHealth, September 1990. |
TABLE 2-2b Investment in Neuroscience Research (in thousands of dollars) by the Alcohol, Drug Abuse, and Mental Health Administration a
TABLE 2-3 U.S. Investment (in thousands of dollars) in Neuroscience and Mental Health Research: Sponsoring Agencies and Foundations a
Unit |
Amount b |
National Institutes of Health |
992,254 |
Alcohol, Drug Abuse, and Mental Health Administration |
275,474 |
Department of Veterans Affairs |
115,369 c |
National Science Foundation |
39,000 |
Office of Naval Research |
22,200 |
Department of Energy |
20,000 |
Office of Scientific Research U.S. Air Force |
15,600 |
Environmental Protection Agency |
3,900 d |
Department of Agriculture |
3,501 |
Centers for Disease Control |
1,204 e |
Foundations |
|
Howard Hughes Medical Institute |
35,000 |
The John D. and Catherine T. MacArthur Foundation |
14,000 |
Pew Charitable Trusts |
6,500 c |
Whitehall Foundation, Inc. |
2,000 |
Total |
1,546,002 |
a Includes basic and clinical, extramural and intramural research only (services excluded). b Figures are for fiscal year 1990 except where otherwise indicated and were obtained through personal communications and, in some cases, annual reports. c Fiscal year 1989 figures. d Figure is for the Neurotoxicology Division of the Health Effects Research Laboratory. e Extramural grants for head injury research. |
form. The incorporation, through a Brain Mapping Initiative, of enabling technologies in the field of neuroscience research is an appropriate strategy to future breakthroughs.
References
Alcohol , Drug Abuse , and Mental Health Administration. 1990. Alcohol and Health (Seventh Special Report to the U.S. Congress). ADAMHA Pub. No. 90-1656. Washington, D.C. : U.S. Department of Health and Human Services, Public Health Service.
Barker, D. , E. Wright, K. Nguyen, L. Cannon , P. Fain, D. Goldgar, D. T. Bishop , J. Carey , B. Baty , J. Kivlin , H. Willard , J. S. Waye , G. Greig , L. Leinwand , Y. Nakamura. P. O'Connell, M. Leppert , J.-M. Lalouel , R. White , and M. Skolnick. 1987. Gene for
von Recklinghausen neurofibromatosis is in the pericentromeric region of chromosome 17. Science 236 : 1100–1102.
Colwell, R. R. , ed. 1989. Biomolecular Data: A Resource in Transition. New York : Oxford University Press.
DeLisi, C. 1988. Computers in molecular biology: Current applications and emerging trends. Science 240 : 47–52.
Gerstein, D. R. , and H. J. Harwood , eds. 1990. Treating Drug Problems , vol. 1. Washington, D.C. : National Academy Press.
Howard Hughes Medical Institute. 1990. Finding the Critical Shapes. Bethesda, Md. :Howard Hughes Medical Institute Office of Communications.
Hubel, D. H. , and T. N. Wiesel. 1979. Brain mechanisms of vision. Scientific American 241(3) : 150–162.
Lewis, R. 1990. A glimpse of neurofibromatosis 1 protein function. Journal of NIH Research 2(Oct.) : 60–64.
McCormick, B. H. , T. A. DeFanti , and M. D. Brown , eds. 1987. Visualization in scientific computing. SIGGRAPH Computer Graphics Newsletter 21(6) : 1–13.
National Advisory Neurological and Communicative Disorders and Stroke Council. 1989. Decade of the Brain: Answers Through Scientific Research. NIH Pub. No. 88-2957. Bethesda, Md. : U.S. Department of Health and Human Services, National Institutes of Health.
National Institute of Neurological Disorders and Stroke. 1989. Profile. Bethesda, Md. : U.S. Department of Health and Human Services, National Institutes of Health.
Regier, D. A., M. E. Farmer , D. S. Rae , B. Z. Locke , S. J. Keith , L. L. Judd , and F. K. Goodwin. 1990. Comorbidity of mental disorders with alcohol and other substance abuse. Journal of the American Medical Association 264(19) : 2511–2518.
Rice, D. P. , S. Kelman , L. S. Miller , and S. Dunmeyer , eds. 1990. The Economic Costs of Alcohol and Drug Abuse and Mental Illness. ADAMHA Pub. No. 90-1694. Washington, D.C. : U.S. Department of Health and Human Services, Alcohol, Drug Abuse, and Mental Health Administration.
Rowbotham, M. C. , and H. L. Fields. 1989. Post-herpetic neuralgia: The relation of pain complaint, sensory disturbance, and skin temperature. Pain.
Sacks, O. 1970. The Man Who Mistook His Wife for a Hat. New York : Harper and Row.
Seizinger, B. R. , G. A. Rouleau , L. J. Ozelius , A. H. Lane , A. G. Faryniarz , M. V. Chao , S. Huson , B. R. Korf , D. M. Parry , M. A. Pericak-Vance , and J. Gusella. 1987. Genetic linkage of von Recklinghausen neurofibromatosis to the nerve growth factor receptor gene. Cell 49(5) : 589–594.
Smith, T. F. 1990. The history of the genetic sequence databases. Genomics 6 : 701-707.
Vela, C. 1990. Overview of U.S. Genome and Selected Scientific Databases. Background paper prepared for the Committee on a National Neural Circuitry Database , Institute of Medicine.