BACKGROUND We examined the neurobiological mechanisms underlying stress susceptibility using structural magnetic resonance imaging and diffusion tensor imaging to determine neuroanatomic differences between stress-susceptible and resilient mice. We also examined synchronized anatomic differences between brain regions to gain insight into the plasticity of neural networks underlying stress susceptibility. METHODS C57BL/6 mice underwent 10 days of social defeat stress and were subsequently tested for social avoidance. For magnetic resonance imaging, brains of stressed (susceptible, n = 11; resilient, n = 8) and control (n = 12) mice were imaged ex vivo at 56 μm resolution using a T2-weighted sequence. We tested for behavior-structure correlations by regressing social avoidance z-scores against local brain volume. For diffusion tensor imaging, brains were scanned with a diffusion-weighted fast spin echo sequence at 78 μm isotropic voxels. Structural covariance was assessed by correlating local volume between brain regions. RESULTS Social avoidance correlated negatively with local volume of the cingulate cortex, nucleus accumbens, thalamus, raphe nuclei, and bed nucleus of the stria terminals. Social avoidance correlated positively with volume of the ventral tegmental area (VTA), habenula, periaqueductal gray, cerebellum, hypothalamus, and hippocampal CA3. Fractional anisotropy was increased in the hypothalamus and hippocampal CA3. We observed synchronized anatomic differences between the VTA and cingulate cortex, hippocampus and VTA, hippocampus and cingulate cortex, and hippocampus and hypothalamus. These correlations revealed different structural covariance between brain regions in susceptible and resilient mice. CONCLUSIONS Stress-integrative brain regions shape the neural architecture underlying individual differences in susceptibility and resilience to chronic stress.

No abstract

Manganese-enhanced magnetic resonance imaging (MEMRI) is a widely used technique in rodent neuroimaging studies. Traditionally, Mn is delivered to animals via a systemic injection; however, this can lead to toxic effects at high doses. Recent studies have shown that subcutaneously implanted mini-osmotic pumps can be used to continuously deliver manganese chloride (MnCl), and that they produce satisfactory contrast while circumventing many of the toxic side effects. However, neither the time-course of signal enhancement nor the effect of continuous Mn delivery on behaviour, particularly learning and memory, have been well-characterized. Here, we investigated the effect of MnCl dose and route of administration on a) spatial learning in the Morris Water Maze and b) tissue signal enhancement in the mouse brain. Even as early as 3 days after pump implantation, infusion of 25-50 mg/kg/day MnCl via osmotic pump produced signal enhancement as good as or better than that achieved 24 h after a single 50 mg/kg intraperitoneal injection. Neither route of delivery nor MnCl dose adversely affected spatial learning and memory on the water maze. However, especially at higher doses, mice receiving MnCl via osmotic pumps developed skin ulceration which limited the imaging window. With these findings, we provide recommendations for route and dose of MnCl to use for different study designs.