Fig. 7. Brain Mg²⁺ supplementation modifies synaptic configuration in the hippocampus.

a Experimental design. Tissue samples were collected from CA1 stratum radiatum (s.r.) of the hippocampus for serial scanning electron microscopy (S.E.M.) and immunohistology. Note these animals were tested for learning and memory by water maze before tissue collection (see Supplementary Fig. 10). b An example of a spiny synapse from serial S.E.M. images at a 70-nm thickness (n = 7 rats, the same as in c–e). c Examples of 3D-reconstructed synapses from rats of control and MgT-treated. Note the difference in PSD area (yellow) and spine head volume (red) between the two groups. Box, the synapse shown in (b). d, e Distribution of PSD area (d) and spine head volume (e) (n = 165, 187 synapses from 3, 4 rats for control and MgT groups; two-sided Kolmogorov–Smirnov tests, **P = 0.0064, ****P < 0.0001). Green lines, distributions of PSD area and spine head volume in young adult animals (data from refs. ^118,119). f Examples of synapses (juxtaposed contacts) labeled by PSD95 and VGLUT1 on 70-nm ultrathin slices. VGLUT1 was used to identify presynaptic contacts. Note the difference in PSD95⁺ puncta between the two groups. g Distribution of [PSD95] in single synaptic contacts from control and MgT-treated rats (n = 438, 682 synapses from 3, 4 rats; two-sided Kolmogorov–Smirnov test, ****P < 0.0001). Source data are provided as a Source Data file.