Search

The subiculum is one of output structures of the hippocampal formation and plays a pivotal role in learning and memory. Because its morphological features are less investigated than those of the hippocampus proper, we explored the internal structure of the rat subiculum using immunohistochemistry. The septal subiculum comprised one region, whereas the temporal subiculum consisted of two subregions, the distal subiculum (Sub1) and proximal subiculum (Sub2). The Sub2 contained four layers: (1) molecular layer, (2) superficial cell layer containing three types of pyramidal neurons immunolabeled for either nitric oxide synthase, Purkinje cell protein 4 (PCP4), or calbindin, (3) middle cell layer where boutons labeled for zinc transporter 3 and those for vesicular glutamate transporter 2 accumulated in the proximal and distal part of the Sub2, respectively, and (4) deep cell layer containing PCP4-positive pyramidal cells, apical dendrites of which showed a characteristic bundling pattern. These features were not observed in either the Sub1 or septal subiculum. There were clear species differences between rats and mice in labeling patterns of pyramidal cells. Morphological differences along the septotemporal axis might be the basis of the functional diversity of the subiculum, such as spatial memory and emotional memory processed at different septotemporal levels.

The subiculum is the output component of the hippocampal formation and holds a key position in the neural circuitry of memory. Previous studies have demonstrated the subiculum's connectivity to other brain areas in detail; however, little is known regarding its internal structure. We investigated the cytoarchitecture of the temporal and mid-septotemporal parts of the subiculum using immunohistochemistry. The border between the CA1 region and subiculum was determined by both cytoarchitecture and zinc transporter 3 (ZnT3)-immunoreactivity (IR), whereas the border between the subiculum and presubiculum (PreS) was partially indicated by glutamate receptor 1 (GluR1)-IR. The subiculum was divided into proximal and distal subfields based on cytoarchitecture and immunohistochemistry for calbindin (CB), nitric oxide synthase (NOS) and Purkinje cell protein 4 (PCP4). The proximal subiculum (defined here as subiculum 2) was composed of five layers: the molecular layer (layer 1), the medium-sized pyramidal cell layer (layer 2) that contained NOS- and PCP4-positive neurons, the large pyramidal cell layer (layer 3) characterized by the accumulation of ZnT3- (more proximally) and vesicular glutamate transporter 2-positive (more distally) boutons, layer 4 containing polymorphic cells, and the deepest layer 5 composed of PCP4-positive cells with long apical dendrites that reached layer 1. The distal subiculum (subiculum 1) consisting of smaller neurons did not show these features. Quantitative analyses of the size and numerical density of somata substantiated this delineation. Both the proximal-distal division and five-layered structure in the subiculum 2 were confirmed throughout the temporal two-thirds of the subiculum. These findings will provide a new structural basis for hippocampal investigations.

The mutation of serine-174 to phenylalanine that causes a defect in the Escherichia coli F 1 -ATPase β-subunit is suppressed by further mutations; Gly-149 to Ser, Ala-295 to Thr, Ala-295 to Pro, or Leu-400 to Gln (Miki, J., Fujiwara, K., Tsuda, M., Tsuchiya, T. and Kanazawa, H. (1990) J. Biol. Chem. 265, 21567–21572). We analyzed the effects of these second site mutations and of a newly identified Asn-158 to Tyr mutation on the activities of the ATPase without the original Ser-174 to Phe mutation. The β-subunit with each amino acid replacement was expressed in the mutant strain JP17, which does not have a β-subunit. Cells transformed with the plasmid carrying Ala-295 to Pro mutation alone did not grow on minimal medium agar supplemented with succinate as the sole carbon source, and showed 3% of the wild-type ATPase activity, suggesting that this mutation caused structural alterations affecting the catalytic function of the enzyme. Conversely transformants with other mutations grew well and had higher ATPase activities, suggesting that these mutations did not cause extensive structural alterations. From the transformants with the plasmid carrying the Ala-295 to Pro mutation, seven revertants capable of cell growth on succinate plates were isolated and reversion mutations were identified at residues 140, 159, 166, 171, 172 and 184 of the β-subunits. The results suggested that Ser-174 and Ala-295 do not necessarily interact directly, but that the regions including these suppression mutation sites close to Ser-174, and Ala-295 interact with each other for the proper functioning of the ATPase. The ternary structure of the region surrounded by the residues which were identified as the reversion mutation sites for Ser-174 to Phe and Ala-295 to Pro mutations is important for the catalytic function of this enzyme.