Cortical interneurons are generated in the ganglionic eminences and migrate through the ventral and dorsal telencephalon before finding their final positions within the cortical plate. During early stages of migration, these cells are present in two well-defined streams within the developing cortex. In an attempt to identify candidate genes which may play a role in interneuron stream specification, we previously carried out a microarray analysis which identified a number of cadherin receptors that were differentially expressed in these streams, including Cadherin-13 (Cdh13). Expression analysis confirmed Cdh13 to be present in the preplate layer at E13.5 and, later in development, in some cortical interneurons and pyramidal cells. Analysis of Cdh13 knockout mice at E18.5, but not at E15.5, showed a reduction in the number of interneurons and late born pyramidal neurons and a concomitant increase in apoptotic cells in the cortex. These observations were confirmed in dissociated cell cultures using overexpression and short interfering RNAs (siRNAs) constructs and dominant negative inhibitory proteins. Our findings identified a novel protective role for Cdh13 in cortical neuron development.

The laying down of memory requires strong stimulation resulting in specific changes in synaptic strength and corresponding changes in size of dendritic spines. Strong stimuli can also be pathological, causing a homeostatic response, depressing and shrinking the synapse to prevent damage from too much Ca2+ influx. But do all types of dendritic spines serve both of these apparently opposite functions? Using confocal microscopy in organotypic slices from mice expressing green fluorescent protein in hippocampal neurones, the size of individual spines along sections of dendrite has been tracked in response to application of tetraethylammonium. This strong stimulus would be expected to cause both a protective homeostatic response and long-term potentiation. We report separation of these functions, with spines of different sizes reacting differently to the same strong stimulus. The immediate shrinkage of large spines suggests a homeostatic protective response during the period of potential danger. In CA1, long-lasting growth of small spines subsequently occurs consolidating long-term potentiation but only after the large spines return to their original size. In contrast, small spines do not change in dentate gyrus where potentiation does not occur. The separation in time of these changes allows clear functional differentiation of spines of different sizes.