Dissociable Networks of the Lateral/Medial Mammillary Body in the Human Brain

Abstract

The mammillary body (MB) has been thought to implement mnemonic functions. Although recent animal studies have revealed dissociable roles of the lateral and medial parts of the MB, the dissociable roles of the lateral/medial MB in the human brain is still unclear. Functional connectivity using resting-state functional magnetic resonance imaging (fMRI) provides a unique opportunity to noninvasively inspect the intricate functional organization of the human MB with a high degree of spatial resolution. The present study divided the human MB into lateral and medial parts and examined their functional connectivity with the hippocampal formation, tegmental nuclei, and anterior thalamus. The subiculum of the hippocampal formation was more strongly connected with the medial part than with the lateral part of the MB, whereas the pre/parasubiculum was more strongly connected with the lateral part than with the medial part of the MB. The dorsal tegmental nucleus was connected more strongly with the lateral part of the MB, whereas the ventral tegmental nucleus showed an opposite pattern. The anterior thalamus was connected more strongly with the medial part of the MB. These results confirm the extant animal literature on the lateral/medial MB and provide evidence on the parallel but dissociable systems involving the MB that ascribe mnemonic and spatial-navigation functions to the medial and lateral MBs, respectively.

Keywords: fornix; hippocampus; mammillotegmental tract; resting-state functional connectivity; tegmentum.

Figure 1

Mammillary body (MB) and regions connected with the MB. The MB receives inputs from the hippocampal formation and tegmental nuclei and sends outputs to the tegmental nuclei and anterior thalamus. By dividing the MB into lateral and medial parts, the functional connectivities with the hippocampal formation, tegmental nuclei, and anterior thalamus were investigated in the present study. Schematic drawings of the subcortical structures are shown in the middle row, and magnetic resonance imaging (MRI) images of the corresponding slices for the dashed lines are shown in the lowest row.

Figure 2

Functional connectivity in the subiculum. (A) Voxel-wise maps of functional connectivity in the ipsilateral hippocampus (seed: the whole MB) shown in the sagittal sections of functional images in one representative subject. The color scale indicates the Gaussian z score of the functional connectivity. The hippocampus and subiculum are delineated by yellow curves. X indicates the X coordinate of the Montreal Neurological Institute (MNI) space. A, anterior; P, posterior; D, dorsal; V, ventral. (B) Voxel-wise maps of differential functional connectivity in the hippocampus (seed: the medial vs. lateral MB). The color scale indicates the Gaussian z score of the differential functional connectivity (hot, medial > lateral; winter, lateral > medial). (C) Gaussian z score averaged across voxels in the left/right subiculum (seed: medial/lateral MB). The error bars indicate the standard error of means across subjects. *P < 0.05, **P < 0.01, paired t-test.

Figure 3

Functional connectivity in the pre/parasubiculum. (A) Voxel-wise maps of functional connectivity in the hippocampus (seed: the whole MB) shown in the coronal sections of one representative subject. The hippocampus, subiculum, and pre/parasubiculum are delineated by yellow curves. Y indicates the Y coordinate of the MNI space. L, left; R, right. (B) Voxel-wise maps of differential functional connectivity in the hippocampus (seed: the medial vs. lateral MB). (C) Gaussian z score averaged across voxels in the left/right pre/parasubiculum (seed: medial/lateral MB). *P < 0.05, paired t-test.

Figure 4

Functional connectivity in the ventral tegmental nucleus. (A) Voxel-wise maps of functional connectivity in the midbrain (seed: the whole MB) shown in the transverse sections of one representative subject. The ventral tegmental nucleus is delineated by a purple curve. Z indicates the Z coordinate of the MNI space. (B) Voxel-wise maps of differential functional connectivity in the midbrain (seed: the medial vs. lateral MB). (C) Gaussian z score averaged across voxels in the left/right ventral tegmental nucleus (seed: medial/lateral MB). *P < 0.05, paired t-test.

Figure 5

Functional connectivity in the dorsal tegmental nucleus. (A) Voxel-wise maps of functional connectivity in the midbrain (seed: the whole MB) shown in the transverse sections of one representative subject. The dorsal tegmental nucleus is delineated by a purple curve. (B) Voxel-wise maps of differential functional connectivity in the midbrain (seed: the medial vs. lateral MB). (C) Gaussian z score averaged across voxels in the left/right dorsal tegmental nucleus (seed: medial/lateral MB). *P < 0.05, **P < 0.01, paired t-test.

Figure 6

Functional connectivity in the anterior thalamus. (A) Voxel-wise maps of functional connectivity in the thalamus (seed: the whole MB) shown in the coronal sections of one representative subject. The whole thalamus and anterior thalamus are delineated by yellow curves. (B) Voxel-wise maps of differential functional connectivity in the thalamus (seed: the medial vs. lateral MB). (C) Gaussian z averaged across voxels in the left/right anterior thalamus (seed: medial/lateral MB). *P < 0.05, paired t-test.