The BDNF Val66Met Polymorphism Affects the Vulnerability of the Brain Structural Network

Abstract

Val66Met, a naturally occurring polymorphism in the human brain-derived neurotrophic factor (BDNF) gene resulting in a valine (Val) to methionine (Met) substitution at codon 66, plays an important role in neuroplasticity. While the effect of the BDNF Val66Met polymorphism on local brain structures has previously been examined, its impact on the configuration of the graph-based white matter structural networks is yet to be investigated. In the current study, we assessed the effect of the BDNF polymorphism on the network properties and robustness of the graph-based white matter structural networks. Graph theory was employed to investigate the structural connectivity derived from white matter tractography in two groups, Val homozygotes (n = 18) and Met-allele carriers (n = 55). Although there were no differences in the global network measures including global efficiency, local efficiency, and modularity between the two genotype groups, we found the effect of the BDNF Val66Met polymorphism on the robustness properties of the white matter structural networks. Specifically, the white matter structural networks of the Met-allele carrier group showed higher vulnerability to targeted removal of central nodes as compared with those of the Val homozygote group. These findings suggest that the central role of the BDNF Val66Met polymorphism in regards to neuroplasticity may be associated with inherent differences in the robustness of the white matter structural network according to the genetic variants. Furthermore, greater susceptibility to brain disorders in Met-allele carriers may be understood as being due to their limited stability in white matter structural connectivity.

Keywords: BDNF Val66Met; diffusion tensor imaging; network resilience; tractography; white matter structural network.

Figure 1

(A) Group-averaged reconstructed white matter structural networks in each group of Val homozygotes (blue) and Met-allele carriers (dark gray) and (B) three-dimensional representations (axial and sagittal views) of white matter network modules of each group. Each node is color-coded by the modular structures. The size of the nodes in (A,B) is in proportion to the number of degrees of each node. The nodes and edges of white matter structural networks were visualized using the BrainNet Viewer (Xia et al., 2013). Abbreviations: Val, valine; Met, methionine.

Figure 2

Network robustness of the white matter structural network in each group of Val homozygotes (blue) and Met-allele carriers (dark gray) in response to targeted node attacks (A) and targeted edge attacks (B). The line graphs indicate changes in the largest connected component size (left panel) and global efficiency (right panel) as a function of nodes or edges removed according to their betweenness centrality in a decreasing order. The bar graphs show the comparisons of AUCs of the largest component size (left panel) or global efficiency (right panel) between Val homozygotes and Met-allele carriers. Asterisks indicate a significant group difference at Bonferroni-corrected p < 0.05. The error bars represent 95% confidence intervals. Abbreviations: Val, valine; Met, methionine; AUC, area under the curve.

Figure 3

Network robustness of the white matter structural network in each group of Val homozygotes (blue) and Met-allele carriers (dark gray) in response to random node attacks (A) and random edge attacks (B). The line graphs indicate changes in the largest connected component size (left panel) and global efficiency (right panel) as a function of randomly removed nodes or edges with 1000 permutations. The bar graphs shows the comparisons of AUCs of the largest connected component size (left panel) or global efficiency (right panel) between Val homozygotes and Met-allele carriers. The error bars represent 95% confidence intervals. Abbreviations: Val, valine; Met, methionine; AUC, area under the curve.