Search

GluR2 is the regulatory subunit in the AMPA family of glutamate receptors (GluRs) in that its presence inhibits calcium flux and dominates the current/ voltage characteristics of AMPA receptors. Studies from other laboratories have shown that GABAergic interneurons have a lower ratio of GluR2/GluR1 mRNA than pyramidal cells as well as possessing AMPA receptors that have a higher relative permeability to calcium. We hypothesized that such differences might be related to differences in the subunit stoichiometry at the AMPA synapses in each cell class, and used a GluR2-specific monoclonal antibody in a double-label immunogold protocol with anti-GABA and anti-CaM kinase II to compare the GluR2 representation at asymmetric synapses in GABA neurons to that of pyramidal cells in rat CA1. Virtually all CA1 pyramidal cells as well as the majority of GABAergic interneurons were GluR2 positive. EM immunogold labeling also showed that GABAergic interneurons had distinctive ultrastructural features and contained GluR2 in both their soma and their dendrites, as did the spines and shafts of pyramidal cells. GluR2 immunoreactivity was frequently preferentially located at asymmetric synapses on both pyramidal cell spines and shafts as well as the dendritic processes and soma of GABAergic interneurons. However, the labeled synapses on GABAergic neurons had a significantly lower number of immunogold particles than those on pyramidal cells. In fact, 90% of the labeled asymmetric synapses on GABAergic cells had one to three gold particles, whereas greater than 70% of the labeled asymmetric synapses on pyramidal cells had four or more gold particles associated with the synapse. These data suggest that while both cell classes contain GluR2, they differ in the relative representation of GluR2 at their AMPA synapses, such that GABAergic neurons might possess AMPA receptors with higher calcium permeability on average than pyramidal cells. Such differences in subunit representation at AMPA-receptor-mediated synapses would not only lead to differences in calcium permeability and functional characteristics across these two cell classes, but might also be relevant to the hippocampal patterns of selective vulnerability with respect to excitotoxicity and neurodegeneration.

Due to its role as the dominant AMPA receptor subunit in respect to regulation of calcium permeability, information on the neuronal localization of GluR2 is of particular importance, yet has been hampered by the lack of a GluR2-specific antibody. Monoclonal antibodies were raised against the putative N-terminal portion (amino acids 175--430) of GluR2, using the fusion protein linked to trpE as an antigen. Western blot analysis and immunocytochemistry of transiently transfected human embryonic kidney 293 cells unambiguously confirmed the specificity of monoclonal antibody 6C4 for GluR2, which did not recognize or cross-react with any other AMPA/Kainate GluR subunits expressed. 6C4 was used in immunohistochemical studies to characterize the regional, cellular, and subcellular distribution of the GluR2 subunit at the light and electron microscopic levels in rat hippocampus and somatosensory cortex and in colocalization studies with the three calcium-binding proteins: parvalbumin, calbindin, and calretinin. GluR2 was widely distributed in both pyramidal cells and interneurons. Asymmetric synapses were labeled on both spines and small dendritic shafts. In contrast to previous reports, our double labeling studies using monoclonal antibody 6C4 with polyclonal antisera against calcium-binding proteins demonstrated that 84--97% of parvalbumin and calbindin-immunoreactive and 45--66% of the calretinin-immunoreactive interneurons in CA1 and somatosensory cortex also contain GluR2. These data have important implications regarding heterogeneity in calcium permeability of AMPA receptors across cell types in neocortex and hippocampus, as well as for differential vulnerability to excitotoxic injury.