Functional Neuroscience of Mood Disorders
Michel Le Moal and Willy Mayo
The evolution of criteria, symptoms and models for diagnosing mood disorders, as reflected in classical twentieth-century textbooks, has been a long process, if we consider Kraepelin's (1921) phenomenological characterizations to constitute a historical landmark. This is especially true for classification and the progressive importance of biological theories. A depressive state is considered to result from either endogenous determinants or a reaction to environmental or life events and a failure to adapt. The evolution from the first edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-I) (American Psychiatric Association, 1952) to the fourth edition (DSM-IV) (American Psychiatric Association, 1994) illustrates, through the push for more reliable and valid diagnoses and for a change to a categorical system (notably for research), profound changes due largely to the new pharmacological thought and its aminergic theories and derivatives. However, it remains clear mat the parallel neurobiological investigation on mood has not played a pivotal role, and that pathophysiological mechanisms have still not been identified, at least in a specific manner, in relation to the categories proposed by the last DSM (Boland and Keller, 1999). Moreover, the relation between the genetic aspects of susceptibility to mood disorders (mainly bipolar) and possible defects at the neuronal levels have not yet been demonstrated. These problems are now at the forefront of psychiatric research. Thanks to advances in statistical analysis combined with powerful new biotechnology methods, hypotheses can now be explored more rapidly. Replicated linkage has, at present, proposed susceptibility loci on chromosomes X, 18 and 21, but a single locus genetic model is in no way responsible for unipolar, bipolar and recurrent disorders (for review, see Sanders et al, 1999), leading to multifactorial vulnerabilities, including environmental factors as well as many genes. Finally, promising data from brain-imaging techniques have opened a new era to help identify circuitries involved in the physiopathology and proposed new neuroanatomies of mood disorders (Manji et al, 2001). However, these investigations raise methodological problems.
THEORETICAL AND METHODOLOGICAL
of Mood Disorders
Basic and experimental neurosciences and clinical neuropsychiatry have joined forces over the past decade to promote the biological foundations of mental illness. The task is so difficult that it can be considered, together with the mind-consciousness problem, as the last frontier in biomedical research and the definitive inclusion of psychiatry in medical sciences. In spite of significant advances that have transformed the clinical perspectives in neurology and related disorders (obsessive-compulsive disorders and Tourette's syndrome are good examples), the growing tendency to 'neurologize' classic psychiatric syndromes, such as the various forms of schizophrenia, has not yet led to universal consensus regarding the aetiology and pathophysiology of such major mental illnesses.
From a neurobiological standpoint, and in the perspective of the neuroanatomical bases of the syndromes, three main general questions remain to be solved. First, when considering a given syndrome or disorder, or a specific mental disease, the symptoms that define the disease still have to be related to a given (or a set of related) 'normal' integrated functions. In other words, a continuum from physiology to pathology and vice versa is expected and, if so, then these functional dimensions should correlate with anatomical loci. The neurobiological interpretations of the symptoms have to be placed within the framework of the structure-function paradigm (Pearlson, 1999), in which modern neurosciences and neuropsychology have made enormous progress, even regarding mental functions and higher cognitive processes. The neuropsychological paradigm first applied to the neurology of cortical functions is now open to psychiatric symptoms. For instance, the disinhibition syndrome characteristic of acute mania will be referred to a similar syndrome sometimes seen after frontal brain injury, and mania will be attributed to a dysfunction of this region.
Second, by the same token, using working hypotheses inspired by experimental medicine, and considering that the concept of experimental psychopathology results directly from the structure-function paradigm, the neurobiologist could create appropriate animal models of the given illness by manipulating either the function or the underlying structure. In this context, the interpretation of experimental conditions is limited by the fact that most of the major human mental disorders do not exist per se in animals. Moreover, two other problems render experimental psychopathology difficult: on the one hand, manipulation of the model to obtain all the basic symptoms or dimensions that could characterize the illness or syndrome is theoretically impossible because they are known mainly by subjective report, and on the other hand, these basic symptoms may also be found in other heterogeneous disorders, because of the difficulty of defining mental illness, as reflected by ever changing concepts in psychiatry, or due to the clinical reality of considerable comorbidities. For instance, many depressed patients display anxiety symptoms, while anxiety and depression are also considered separately in DSM-IV.
Third, the pharmacological paradigm is not totally satisfying. It is based on a well-known working hypothesis: (1) the active drug acts on neuronal systems that are defective and part of the pathophysiology of the disease; (2) it is of major interest to discover the site of action of therapeutic drugs, and these sites or structures have generally been characterized as being the specific