The BDNFVal66Met SNP modulates the association between beta-amyloid and hippocampal disconnection in Alzheimer's disease

Abstract

In Alzheimer's disease (AD), a single-nucleotide polymorphism in the gene encoding brain-derived neurotrophic factor (BDNF_Val66Met) is associated with worse impact of primary AD pathology (beta-amyloid, Aβ) on neurodegeneration and cognitive decline, rendering BDNF_Val66Met an important modulating factor of cognitive impairment in AD. However, the effect of BDNF_Val66Met on functional networks that may underlie cognitive impairment in AD is poorly understood. Using a cross-validation approach, we first explored in subjects with autosomal dominant AD (ADAD) from the Dominantly Inherited Alzheimer Network (DIAN) the effect of BDNF_Val66Met on resting-state fMRI assessed functional networks. In seed-based connectivity analysis of six major large-scale networks, we found a stronger decrease of hippocampus (seed) to medial-frontal connectivity in the BDNF_Val66Met carriers compared to BDNF_Val homozogytes. BDNF_Val66Met was not associated with connectivity in any other networks. Next, we tested whether the finding of more pronounced decrease in hippocampal-medial-frontal connectivity in BDNF_Val66Met could be also found in elderly subjects with sporadically occurring Aβ, including a group with subjective cognitive decline (N = 149, FACEHBI study) and a group ranging from preclinical to AD dementia (N = 114, DELCODE study). In both of these independently recruited groups, BDNF_Val66Met was associated with a stronger effect of more abnormal Aβ-levels (assessed by biofluid-assay or amyloid-PET) on hippocampal-medial-frontal connectivity decreases, controlled for hippocampus volume and other confounds. Lower hippocampal-medial-frontal connectivity was associated with lower global cognitive performance in the DIAN and DELCODE studies. Together these results suggest that BDNF_Val66Met is selectively associated with a higher vulnerability of hippocampus-frontal connectivity to primary AD pathology, resulting in greater AD-related cognitive impairment.

Fig. 1

Surface renderings of significant seed-based connectivity (p < 0.001, FWE cluster corrected at p < 0.05) in the DIAN discovery sample for the primary hippocampal seed ROIs and the secondary seed ROIs to derive connectivity for the control network (CON), dorsal attention network (DAN), default mode network (DMN), and salience network (SAL)

Fig. 2

Results of the voxel-wise interaction analysis of EYO × BDNF_Val66Met on hippocampal connectivity in the DIAN-MC subjects for the right (a, b) and left (e, f) hippocampus seed ROI (i.e. discovery). In panels a and e, red areas correspond to regions showing significant seed-based hippocampal connectivity, while purple and blue regions indicate the boundaries of the significant EYO × BDNF_Val66Met interaction. Validation analyses of the BDNF_Val66Met × Aβ interaction in DELCODE and FACEHBI using the DIAN-MC derived medial-frontal ROI are shown in panels c and g for DELCODE and d and h for FACEHBI. DIAN-MC = mutation carriers with autosomal dominant AD from DIAN

Fig. 3

Scatterplots illustrating the interaction between right (a–d) and left (e–h) hippocampus to medial-frontal connectivity and AD severity (i.e. EYO or CSF Aβ42/40 ratio) on cognition. Color groupings were based on median split and are for illustrational purposes only, since the underlying linear mixed models were computed using continuous measures. DIAN-MC = mutation carriers with autosomal dominant AD from DIAN