Search

Characterizing age- and risk-related hippocampal vulnerabilities may inform about the neural underpinnings of cognitive decline. We studied the impact of three risk-factors, Apolipoprotein (APOE)-ε4, a family history of dementia, and central obesity, on CA1, CA2/3, dentate gyrus (DG) and subiculum in 158 cognitively healthy adults (38-71 years). Subfields were labelled with the Automatic Segmentation of Hippocampal Subfields (ASHS) and FreeSurfer (version 6) protocols. Volumetric and microstructural measurements from quantitative magnetization transfer and Neurite Orientation Density and Dispersion Imaging were extracted for each subfield and reduced to three principal components capturing apparent myelin/neurite packing, size/complexity, and metabolism. Aging was associated with an inverse U-shaped curve on myelin/neurite packing and affected all subfields. Obesity led to reductions in myelin/neurite packing and size/complexity regardless of APOE and FH status. However, amongst individuals with a healthy Waist-Hip-Ratio, APOE ε4 carriers showed lower size/complexity than non-carriers. Protocol type did not affect this risk pattern. These findings provide novel evidence for interactive effects between APOE and central obesity on the hippocampal formation of cognitively healthy adults.HighlightsAge-related inverted U-shaped curve of hippocampal myelin/neurite packingObesity-related reductions of hippocampal myelin/neurite packing and size/complexityAPOE modifies the effects of obesity on hippocampal size/complexityAge-related slowing of spatial navigationNo APOE, family history, or obesity effects on cognition

Invasive tract-tracing studies in rodents implicate a direct connection between the subiculum and bed nucleus of the stria terminalis (BNST) as a key component of neural pathways mediating hippocampal regulation of the Hypothalamic-Pituitary-Adrenal (HPA) axis. A clear characterisation of the connections linking the subiculum and BNST in humans and non-human primates is lacking. To address this, we first delineated the projections from the subiculum to the BNST using anterograde tracers injected into macaque monkeys, revealing evidence for a monosynaptic subiculum-BNST projection involving the fornix. Second, we used in vivo diffusion MRI tractography in macaques and humans to demonstrate substantial subiculum complex connectivity to the BNST in both species. This connection was primarily mediated through the fornix, with additional connectivity via the amygdala, consistent with rodent anatomy. Third, utilising the twin-based nature of our human sample, we found that microstructural properties of these tracts are moderately heritable (h2 ∼ 0.5). In a final analysis, we found no evidence of any significant association between subiculum complex-BNST tract microstructure and indices of perceived stress/dispositional negativity and alcohol use, derived from principal component analysis decomposition of self-report data. We did, however, find subiculum complex-BNST tract microstructure associations with BMI, age, and sex. Our findings address a key translational gap in our knowledge of the neurocircuitry regulating stress.

Retrosplenial cortex contains two principal subdivisions, area 29 (granular) and area 30 (dysgranular). Their respective anatomical connections in the rat brain reveal that area 29 is the primary recipient of hippocampal and parahippocampal spatial and contextual information while area 30 is the primary interactor with current visual information. Lesion studies and measures of neuronal activity in rodents indicate that retrosplenial cortex helps to integrate space from different perspectives, e.g., egocentric and allocentric, providing landmark and heading cues for navigation and spatial learning. It provides a repository of scene information that, over time, becomes increasingly independent of the hippocampus. These processes, reflect the interactive actions between areas 29 and 30, along with their convergent influences on cortical and thalamic targets. Consequently, despite their differences, both areas 29 and 30 are necessary for an array of spatial and learning problems.