Anhedonia has been one of the two key diagnostic criteria for major depressive disorder (MDD) since the publication of The Diagnostic and Statistical Manual of Mental Disorders, Third Edition, yet little is known about its neural substrates. Neuroimaging studies have identified numerous brain regions that are thought to be involved with MDD. Most studies dealt with MDD as a syndromal entity, and not surprisingly, yielded quite variable results with respect to the areas of the brain identified, the nature of the functional changes (i.e., decrease or increase in activities), lateralization, and correlation with clinical symptoms. Clinical heterogeneity and lack of symptom-specific targets are presumably among the factors contributing to the variability. The hypothesis that a functionally impaired mesolimbic dopaminergic pathway may comprise a part of neural substrate underlying core MDD symptoms of anhedonia and loss of motivation has been proposed. Nevertheless, the roles of brain reward mechanisms in mediating anhedonia in MDD remain unclear. Availability of appropriate experimental paradigms that can be used empirically to measure anhedonia is a prerequisite to test such a hypothesis. Recent studies using monetary incentive paradigms coupled with neuroimaging techniques have identified hemodynamic responses in structures that serve as part of the mesolimbic dopaminergic pathway during processing rewards in healthy humans. We hypothesize that anhedonia in MDD is associated with impairment of brain reward mechanisms such that dysfunction of the orbital and ventromedial frontal cortices involved in the impaired hedonic attribution capacity, while dysfunction of the ventral striatum area that contains the nucleus accumbens is involved with the reduced or lack of reactivity to rewarding environmental stimuli in patients with MDD. Our hypothesis is that specific neural substrates are linked to the two psychiatric components of anhedonia, i.e., loss of interest and lack of reactivity, as defined in the diagnostic criteria for MDD. We plan to test our hypothesis by using empirical measurement of reward responses in MDD patients with and without significant anhedonia using a monetary incentive event-related functional magnetic resonance imaging task recently developed at NIH. We expect to find reduced activation of the ventral striatum, orbital and ventromedial frontal cortices in response to monetary incentive stimuli in MDD patients with significant anhedonia relative to MDD patients without anhedonia and healthy control. The outcome of the proposed work may provide clues for diagnosis, classification, and treatment of MDD, and may also yield leads for identifying the fundamental neural mechanisms underlying anhedonia in other disabling psychiatric conditions such as schizophrenia and addiction. During the past year, PET and fMRI data from depressed and healthy control subjects have been acquired, analyzed, and reported at several international scientific meetings. The behavioral data from these analyses show significant differences in the modulation of performance by incentive magnitude between the depressives and controls. The neural basis for these differences were demonstrated by analysis of the fMRI data, and showed abnormal activation of several regions of the striatum (caudate, nucleus accumbens) and orbitofrontal corex in MDD. Two manuscripts are in preparation to describe these findings. In addition, two new experimental tasks were added, and new pilot data was obtained in 16 healthy control subjects to demonstrate that these tasks were satisfactory. One involves a reward conditioning task and the other is a working memory task in which motivation and attention are differentially modulated by intermittently making monetary reward contingent on the behavioral performance. The initial analyses of the fMRI and neurocognitive data suggested these tasks were effectively tapping into the neural and systems of interest and influencing motivated behavior, so the studies in depressed patients have been initiated. These studies are nearly completed and analyzed, and papers describing the results are in preparation for publication. Furhtemore, PET studies are being conducted using PET and C-11raclopride to investigate endogenous dopamine release during naturally rewarding tasks. We have developed for the first time a method for measuring reward-related dopamine release in striatal regions such as the accumbens area in a manner that is not confounded by activation-associated changes in local blood flow and also is controlled for the motor aspects. We have run 30 healthy volunteers in this protocol and the results show that the amount of dopamine release in the ventral striatum, an area known to play a major role in reward processing in rats and monkeys, is strongly lateralized to the right side in humans. A manuscript is in preparation describing these results and methods. This is now being extended into depressed patients to test the hypothesis that reward-related dopamine release is blunted in depressed patients, and 17 unipolar depressed subjects and 6 bipolar depressed subjects have been imaged thus far. In the unipolar depressives the right ventral striatum showed an abnormal attenuation of dopamine release during reward processing. This observation may account for why depressed patients fail to find reward or pleasure from their activities, relationships or work. More recently, we used PET imaging and C-11NNC112 to image dopamine type 1 (D1) receptors in the brain in patients with depression and healthy controls. In the left middle caudate, the same region where the amount of dopamine release was abnormally reduced in depression, the D1 receptor binding was abnormally decreased in the depressed subjects. The extent to which D1 binding was diminished was correlated with the extent of anhedonia (inability to experience pleasure) that was reported by the patients. This finding thus converges with the other types of imaging data described above to link abnormal processing in multiple aspects of the dopaminergic system to the reason why depressed patients fail to find joy and pleasure in life. Finally, we completed a study of appetitive and aversive learning in patients with depression using the behavior and fMRI. We found that depressed subjects do not acquire a conditioned preference for a rewarding stimulus, but instead acquire one for an aversive stimulus. This was accompanied by abnormal responses in the amygdala and orbitofrontal cortex, which in animals have been shown to have this type of abnormal response once they acquire lesions of the orbitofrontal cortex. This research shows another example of emotional processing biases which occur in depression, and also show the functional anatomical correlates of how this processing goes awry in depression.
