The brain-derived neurotrophic factor Val66Met polymorphism is associated with reduced functional magnetic resonance imaging activity in the hippocampus and increased use of caudate nucleus-dependent strategies in a human virtual navigation task

Abstract

Multiple memory systems are involved in parallel processing of spatial information during navigation. A series of studies have distinguished between hippocampus-dependent 'spatial' navigation, which relies on knowledge of the relationship between landmarks in one's environment to build a cognitive map, and habit-based 'response' learning, which requires the memorization of a series of actions and is mediated by the caudate nucleus. Studies have demonstrated that people spontaneously use one of these two alternative navigational strategies with almost equal frequency to solve a given navigation task, and that strategy correlates with functional magnetic resonance imaging (fMRI) activity and grey matter density. Although there is evidence for experience modulating grey matter in the hippocampus, genetic contributions may also play an important role in the hippocampus and caudate nucleus. Recently, the Val66Met polymorphism of the brain-derived neurotrophic factor (BDNF) gene has emerged as a possible inhibitor of hippocampal function. We have investigated the role of the BDNF Val66Met polymorphism on virtual navigation behaviour and brain activation during an fMRI navigation task. Our results demonstrate a genetic contribution to spontaneous strategies, where 'Met' carriers use a response strategy more frequently than individuals homozygous for the 'Val' allele. Additionally, we found increased hippocampal activation in the Val group relative to the Met group during performance of a virtual navigation task. Our results support the idea that the BDNF gene with the Val66Met polymorphism is a novel candidate gene involved in determining spontaneous strategies during navigation behaviour.

Fig. 1

Visual depiction of the 4/8VM environment.

Fig. 2

Visual depiction of the CSDLT environment. (A) Participants’ view of the environment, showing one of six pairs of pathways. (B) Overhead view of the environment, showing an enriched distal landscape.

Fig. 3

Schematic representation of the CSDLT. Encoding phase: six pairs of pathways are presented individually and repeatedly in a pseudo-random order until participants learn to a criterion of 11/12 correct choices. Test phase recombined condition: reward contingency remains the same, but pairs are recombined into new combinations. Participants complete two trials of four recombined pairs each. Test phase all-open condition: pathways are no longer divided into pairs, and participants have access to the entire environment. Participants are told to collect the objects from the six rewarded pathways while avoiding empty pathways. Each of the letters A, B, C, D, E and F represent a pair of arms presented simultaneously in Stage 1. The letter combinations AB, CD, EF and AF represent the newly recombined pairs of arms in Stage 2. For example, AB represents the combination of one arm from pair A and one arm from pair B.

Fig. 4

Comparison of initial strategy use on the 4/8VM grouped by number of Met alleles (0 = Val/Val; 1 = Val/Met; 2 = Met/Met). The division is shown by the percentage of participants within a given genotype group. (*) A significant association was observed between the number of Met alleles and initial 4/8VM spontaneous strategy (linear-by-linear association, χ² = 4.203, P < 0.04), with the proportion of participants using a non-hippocampal response strategy increasing with the number of Met alleles.

Fig. 5

fMRI activity during the early encoding phase (experimental trial 1). Statistical parametric maps showing engagement of the hippocampus in the Val group during early learning (experimental trial 1). The t-statistic maps are superimposed on the anatomical average of all participants and displayed in the axial, sagittal and coronal planes. Cross-hairs are centred on the voxel with the highest BOLD response activity in the hippocampus for the Val group (x = 23, y = −6, z = −27; t = 4.05). No other region of the brain crossed the threshold for significance corrected for multiple comparisons. The colour bars illustrate the range of t-statistical values shown. For interpretation of references to color in the figure legend, please refer to the Web version of this article.

Fig. 6

fMRI activity during the late encoding phase (the last two experimental trials of the encoding phase) of the Val > Met contrast. Statistical parametric maps showing (A) increased engagement of the caudate nucleus in the Met group as compared with the Val group during late learning (last two experimental trials before the recombined condition) and (B) increased engagement of the caudate nucleus in the Met group as compared with the Val group during the all-open test phase condition. The t-statistic maps are superimposed on the anatomical average of all participants and displayed in the axial, sagittal and coronal planes. Cross-hairs are centred on the voxel with the highest BOLD response activity in the caudate nucleus for the Met group (x = 22, y = −18, z = 26; t = 4.23). In the late learning contrast (A), the cross-hairs are centred on the voxel with the greatest degree of activation in the caudate nucleus (x = 20, y = −19, z = 24; t = −3.14). In the all-open contrast (B), the cross-hairs are centred on the voxel with the greatest degree of activation in the caudate nucleus (x = −6, y = 8, z = 16; t = −3.30). No other region of the brain crossed the threshold for significance corrected for multiple comparisons. The colour bars illustrate the range of t-statistical values shown. For interpretation of references to color in the figure legend, please refer to the Web version of this article.