One can describe the influence of antidepressant medications either by the mechanism of their action at the molecular level, or by the resulting large-scale changes in brain activity observed in treated subjects. At the moment we have a robust picture of both of these, but little sense of how one leads to the other, nor their precise relationship to the psychological effects of treatment.
The most popular type of antidepressant today is the selective serotonin reuptake inhibitor (SSRI), such as Prozac, Zoloft, or Lexapro, introduced in 1987. SSRIs increase synapse concentrations of the neurotransmitter serotonin by binding to and thus blocking serotonin transporters – proteins which normally recycle serotonin molecules back into the neuron that released them.
The relationship between serotonin levels and mood is by no means straightforward, nor well understood; for instance, illegal drugs such as MDMA (ecstasy), psylocibin (mushrooms), and LSD all influence serotonin receptors or transporters, but each have different effects on cognition, suggesting against a simplistic association between serotonin levels and positive mood.
Nevertheless, major depression has been very consistently associated with dysfunction in the serotonin system: depressed subjects have fewer serotonin receptors and transporters (assessed using PET imaging) and decreased levels of tryptophan, a metabolic precursor to serotonin, in cerebrospinal fluid.
Tryptophan depletion consistently induces depression in subjects with high familial risk for or past episodes of depression, but not in normal subjects, and SSRIs do not have a consistent influence on the mood of subjects without depression. Thus, it is likely that SSRIs work by remedying a specific dysfunction in the serotonin system beyond simply upping serotonin levels.
The past fifteen years have witnessed a wealth of studies using neuroimaging techniques such as functional Magnetic Resonance Imaging (fMRI) and positron emission tomography (PET) to measure regional changes in brain activity associated with depression and its treatment.
Depression has been repeatedly associated with increased baseline activity in areas involved in the processing of emotion: the amygdala, subgenual anterior cingular cortex, ventromedial prefrontal cortex, and insula.
Also, depressed subjects exhibit increases in amygdala activation when presented with sad or fearful faces (whether these are consciously or subliminally presented). A similar pattern of changes in activation is observed in subjects experiencing transient mood shifts, such as watching a heartwrenching movie clip.
Treatment with serotonin consistently shifts activation patterns in the opposite direction, decreasing baseline activity in the aforementioned areas and reducing the evoked amygdala response to emotional faces. Further work is needed to detail how SSRIs effect this specific shift in large-scale brain activation patterns, and how exactly this shift corresponds to an alleviation of the symptoms of depression.