Sophie Scotto-Lomassese, Postdoctoral Fellow (until August 2006).
Adult neurogenesis is a widespread phenomenon in vertebrates. In songbirds, new neurons are produced in the ventricular zone (VZ) and migrate throughout the telencephalon. Some of them are incorporated into existing network of some nuclei that belong to the song system, the HVC and the Area X, where they replace neurons that have died. The HVC is composed of 3 populations of neurons: 1) the projection neurons HVCRA that controls the production of learned song, 2) the projection neurons HVCX that are involved in the auditory dependant-song plasticity, 3) interneurons (HVCIN). Interestingly, these 3 populations of HVC neurons can be observed packed tightly together, with extensive soma-soma appositions, forming cell clusters. Moreover, data obtained in canaries showed that half of the newly arriving neurons are recruited into these clusters.
We are interested to study the relative contribution of different HVC cell types toward these clusters and whether recruitment of new neurons into existing functional HVC network depends on correct ‘cluster’ association with particular neuron types. Could the clusters constitute functional units where immature neurons are instructed, perhaps through gap junctions? These questions are addressed using immunocytochemistry (cell birth and cell phenotype markers) in conjunction with stereotaxic injections of retrograde tracers into RA and Area X nuclei.
We first determined which population(s) in the HVC is renewed in adults. New HVC neurons were initially inferred to be HVCIN, because they lacked retrograde labelling from the HVC’s targets. But later studies using different tracers demonstrated that HVCRA are replaced but HVCX are not. We found that none of the new neurons express calbindin, parvalbumin and calretinin that identify the different HVCIN subpopulations. Moreover, we verified that all the HVC-IN express at least one of these calcium-binding proteins. Together, this demonstrates that HVCIN are not replaced in adult [Scotto-Lomassese et al., 2007].
Cell clustering is an important phenomenon in HVC as 80% of the HVC neurons have one or more partners. It involves all 3 the neuron types, but not with the same frequency: HVCXare over-represented in cluster (98.5%) as compared to HVCRA (79.2%) and HVCIN (75.6%). The cluster distribution is not randomly distributed but those composed of two or three neurons were observed most often. Interestingly, one month after their birth, 70% of the new neurons are already found in clusters. Moreover, these newly arriving cells do not incorporate into the most frequently encountered cluster types but apparently associate preferentially with the relatively rare single neurons or join clusters of two cells. In fact, new neurons seem to participate in maintenance of the overall cluster distribution, suggesting the existence of a mechanism that guides new cells toward the right place. We also analyzed the phenotype of the cell in direct contact with a new neuron and found that they are also not distributed according to frequency of occurrence of the different cell classes. Together, these results suggest that integration of new neurons into the HVC is not random and that clusters might participate to the mechanisms controlling the incorporation of the new neurons into the existing network. Ultrastructural studies are currently underway to clarify the presence of gap junctions between new and already existing neurons.