Local calcium release in dendritic spines required for long-term synaptic depression.

We have used rats and mice with mutations in myosin-Va to evaluate the range and function of IP3-mediated Ca2+ signaling in dendritic spines. In these mutants, the endoplasmic reticulum and its attendant IP3 receptors do not enter the postsynaptic spines of parallel fiber synapses on cerebellar Purkinje cells. Long-term synaptic depression (LTD) is absent at the parallel fiber synapses of the mutants, even though the structure and function of these synapses otherwise appear normal. This loss of LTD is associated with selective changes in IP3-mediated Ca2+ signaling in spines and can be rescued by photolysis of a caged Ca2+ compound. Our results reveal that IP3 must release Ca2+ locally in the dendritic spines to produce LTD and indicate that one function of dendritic spines is to target IP3-mediated Ca2+ release to the proper subcellular domain.

Local calcium signalling by inositol-1,4,5-trisphosphate in Purkinje cell dendrites.

The second messenger inositol-1,4,5-trisphosphate (InsP3) releases Ca2+ from intracellular Ca2+ stores by activating specific receptors on the membranes of these stores. In many cells, InsP3 is a global signalling molecule that liberates Ca2+ throughout the cytoplasm. However, in neurons the situation might be different, because synaptic activity may produce InsP3 at discrete locations. Here we characterize InsP3 signalling in postsynaptic cerebellar Purkinje neurons, which have a high level of InsP3 receptors. We find that repetitive activation of the synapse between parallel fibres and Purkinje cells causes InsP3-mediated Ca2+ release in the Purkinje cells. This Ca2+ release is restricted to individual postsynaptic spines, where both metabotropic glutamate receptors and InsP3 receptors are located, or to multiple spines and adjacent dendritic shafts. Focal photolysis of caged InsP3 in Purkinje cell dendrites also produces Ca2+ signals that spread only a few micrometres from the site of InsP3 production. Uncaged InsP3 produces a long-lasting depression of parallel-fibre synaptic transmission that is limited to synapses where the Ca2+ concentration is raised. Thus, in Purkinje cells InP3 acts within a restricted spatial range that allows it to regulate the function of local groups of parallel-fibre synapses.

Local dendritic Ca2+ signaling induces cerebellar long-term depression.

The coordinated activity of large numbers of adjacent parallel fiber synapses elevate calcium concentration locally in small regions of Purkinje cell dendrites. Such activity has also been reported to produce long-term depression of parallel fiber synaptic transmission. We have examined the relationship between these two events by combining patch clamp measurements of parallel fiber synaptic transmission with confocal microscopic imaging of the local calcium signals. We find that patterns of parallel fiber activity capable of evoking long-term depression invariably cause increases in Purkinje cell calcium concentration that are very spatially restricted. These results suggest that one function of the local dendritic calcium signals is to induce long-term depression of parallel fiber synapses.

Calcium as a coagonist of inositol 1,4,5-trisphosphate-induced calcium release.

Inositol 1,4,5-trisphosphate (IP3)-induced calcium release from intracellular stores is a regulator of cytosolic-free calcium levels. The subsecond kinetics and regulation of IP3-induced calcium-45 release from synaptosome-derived microsomal vesicles were resolved by rapid superfusion. Extravesicular calcium acted as a coagonist, potentiating the transient IP3-induced release of calcium-45. Thus, rapid elevation of cytosolic calcium levels may trigger IP3-induced calcium release in vivo. Extravesicular calcium also produced a more slowly developing, reversible inhibition of IP3-induced calcium-45 release. Sequential positive and negative feedback regulation by calcium of IP3-induced calcium release may contribute to transients and oscillations of cytosolic-free calcium in vivo.

Expression of the Alzheimer amyloid precursor gene transcripts in the human brain.

An alternate form of the Alzheimer amyloid protein precursor mRNA that encodes a protease inhibitor domain has recently been reported. Oligonucleotide probes that differentiate between the two mRNAs are used to describe the expression of each amyloid precursor transcript in the human brain. RNA blot analyses show that one of the mRNAs is expressed selectively in the nervous system, that the two messages display different regional distributions in the adult human brain, and that the expression of the two mRNAs is differentially affected in Down's syndrome brain and in Alzheimer's disease frontal cortex. In situ hybridization shows that the two transcripts display the same laminar distribution in the adult cortex but that the transcripts differ significantly in their levels of expression in pyramidal cells of the hippocampus.

Growth-associated protein GAP-43 is expressed selectively in associative regions of the adult human brain.

GAP-43 is a neuron-specific phosphoprotein that has been linked with the development and functional modulation of synaptic relationships. cDNAs for the human GAP-43 gene were used to reveal high overall levels of GAP-43 mRNA in a number of integrative areas of the neocortex, but low levels in cortical areas involved in the initial processing of sensory information, in several brainstem structures, and in caudate-putamen. Neurons expressing highest levels of GAP-43 mRNA were found by in situ hybridization to be concentrated in layer 2 of association cortex and in hippocampal pyramidal cells. Control studies showed that several other RNAs had regional distributions that were different from GAP-43, although the mRNA encoding the precursor of the Alzheimer amyloid beta protein followed a similar pattern of expression. These results suggest that a restricted subset of cortical and hippocampal neurons may be specialized for synaptic remodeling and might play a role in information storage in the human brain.

MAP2 and tau segregate into dendritic and axonal domains after the elaboration of morphologically distinct neurites: an immunocytochemical study of cultured rat cerebrum.

We sought to determine whether the strict segregation of MAP2 and tau into somatodendritic and axonal compartments in situ was maintained in dissociated neuronal cultures of the rat cerebrum. Cultures grown under serum-free conditions were immunolabeled with monoclonal antibodies specific for MAP2 and tau. At 14 d after plating, a clear distinction between MAP2- and tau-immunoreactive neurites was apparent. MAP2-immunoreactive neurites were relatively short, thick, tapering, and branched. Tau-immunoreactive neurites formed a crisscrossing meshwork of long, fine-caliber neurites, which, in more densely plated cultures, had a tendency to form thick, ropelike fascicles. Unlike the MAP2 pattern, tau antibodies labeled somata only lightly. Since distinct populations of neurites were labeled with the 2 antibodies, we sought to observe the development of the topographically distinct compartments by double-labeled immunocytochemistry with both polyclonal and monoclonal antibodies to MAP2 and tau. Cells observed within the first 8 hr after plating demonstrated equally intense MAP2 and tau immunoreactivity in a coextensive distribution throughout the cell body and initial neurites. By 16 hr, some neurites began to assume dendritic and axonal features; however, many such processes contained reaction product for both MAP2 and tau. Beginning at this time, neurites that appeared axonal showed a progressively weaker reaction with MAP2 antibodies, and neurites that appeared dendritic showed a progressively weaker reaction with tau antibodies. In most neurites the diminution appeared to occur uniformly over the entire extent of the neurite. During this transformation period there were occasional axon-like neurites that contained MAP2 immunoreactivity proximally, while tau immunoreactivity extended over the entire length of the neurite. We conclude that neurons in culture are able to compartmentalize MAP2 and tau into their appropriate processes and only attain an apparently homogeneous population of one of these MAPs after the neuron has assumed dendritic and axonal features. The analysis also lends indirect support to the hypothesis that microtubule-associated proteins (MAPs) form this association at the distal extent of the growing neurite.