Depression-associated reductions in putaminal volume are accompanied by a shift from matrix-like to striosome-like structural connectivity.

The striatum is among the most studied brain regions in Major Depressive Disorder (MDD) due to its involvement in reward, motivation, and executive functions. The striatum is comprised of interdigitated tissue compartments (matrix and striosome) that have distinct connectivity, pharmacology, and susceptibility to neuropsychiatric diseases. Each compartment is embedded in distinct functional networks, and functional networks that are abnormal in MDD (eg, the default mode and salience networks) selectively couple with the striosome. We hypothesized that imbalances in function between the striatal compartments could correlate with MDD. Historically, assessing compartment-specific involvement in MDD required histological staining of postmortem tissue, precluding or large-group analyses. In a cohort of 266 subjects (MDD and matched healthy controls), we used probabilistic diffusion tractography (connectivity-based parcellation) to identify striatal voxels with matrix-like and striosome-like patterns of connectivity. These compartment-like voxels recapitulated the anatomic features of matrix and striosome described in prior histologic assessments: their relative abundance, location biases within the striatum, somatotopic organization, and distinct patterns of structural connectivity. Prior studies found that decreased putaminal volume was associated with increased risk of developing MDD. We found decreased putaminal volume in MDD, accompanied by a shift from matrix-like to striosome-like volume. We observed an opposing shift in the caudate, from striosome-like to matrix-like volume, that correlated with the severity of MDD symptoms. Our findings suggest that MDD-associated decreased putaminal volume correlates with a shift from matrix-based functional networks toward striosome-dominated patterns in the putamen. Abnormalities of compartment function in the putamen may be a neuroanatomical correlate of the clinical features of MDD.