Cux1 Enables Interhemispheric Connections of Layer II/III Neurons by Regulating Kv1-Dependent Firing.

Neuronal subtype-specific transcription factors (TFs) instruct key features of neuronal function and connectivity. Activity-dependent mechanisms also contribute to wiring and circuit assembly, but whether and how they relate to TF-directed neuronal differentiation is poorly investigated. Here we demonstrate that the TF Cux1 controls the formation of the layer II/III corpus callosum (CC) projections through the developmental transcriptional regulation of Kv1 voltage-dependent potassium channels and the resulting postnatal switch to a Kv1-dependent firing mode. Loss of Cux1 function led to a decrease in the expression of Kv1 transcripts, aberrant firing responses, and selective loss of CC contralateral innervation. Firing and innervation were rescued by re-expression of Kv1 or postnatal reactivation of Cux1. Knocking down Kv1 mimicked Cux1-mediated CC axonal loss. These findings reveal that activity-dependent processes are central bona fide components of neuronal TF-differentiation programs and establish the importance of intrinsic firing modes in circuit assembly within the neocortex.

Enrichment of conserved synaptic activity-responsive element in neuronal genes predicts a coordinated response of MEF2, CREB and SRF.

A unique synaptic activity-responsive element (SARE) sequence, composed of the consensus binding sites for SRF, MEF2 and CREB, is necessary for control of transcriptional upregulation of the Arc gene in response to synaptic activity. We hypothesize that this sequence is a broad mechanism that regulates gene expression in response to synaptic activation and during plasticity; and that analysis of SARE-containing genes could identify molecular mechanisms involved in brain disorders. To search for conserved SARE sequences in the mammalian genome, we used the SynoR in silico tool, and found the SARE cluster predominantly in the regulatory regions of genes expressed specifically in the nervous system; most were related to neural development and homeostatic maintenance. Two of these SARE sequences were tested in luciferase assays and proved to promote transcription in response to neuronal activation. Supporting the predictive capacity of our candidate list, up-regulation of several SARE containing genes in response to neuronal activity was validated using external data and also experimentally using primary cortical neurons and quantitative real time RT-PCR. The list of SARE-containing genes includes several linked to mental retardation and cognitive disorders, and is significantly enriched in genes that encode mRNA targeted by FMRP (fragile X mental retardation protein). Our study thus supports the idea that SARE sequences are relevant transcriptional regulatory elements that participate in plasticity. In addition, it offers a comprehensive view of how activity-responsive transcription factors coordinate their actions and increase the selectivity of their targets. Our data suggest that analysis of SARE-containing genes will reveal yet-undescribed pathways of synaptic plasticity and additional candidate genes disrupted in mental disease.

Pαx6 expression in postmitotic neurons mediates the growth of axons in response to SFRP1.

During development, the mechanisms that specify neuronal subclasses are coupled to those that determine their axonal response to guidance cues. Pax6 is a homedomain transcription factor required for the specification of a variety of neural precursors. After cell cycle exit, Pax6 expression is often shut down in the precursor progeny and most postmitotic neurons no longer express detectable levels of the protein. There are however exceptions and high Pax6 protein levels are found, for example, in postmitotic retinal ganglion cells (RGCs), dopaminergic neurons of the olfactory bulb and the limbic system in the telencephalon. The function of Pax6 in these differentiating neurons remains mostly elusive. Here, we demonstrate that Pax6 mediates the response of growing axons to SFRP1, a secreted molecule expressed in several Pax6-positive forebrain territories. Forced expression of Pax6 in cultured postmitotic cortical neurons, which do not normally express Pax6, was sufficient to increment axonal length. Growth was blocked by the addition of anti-SFRP1 antibodies, whereas exogenously added SFRP1 increased axonal growth of Pax6-transfected neurons but not that of control or untransfected cortical neurons. In the reverse scenario, shRNA-mediated knock-down of Pax6 in mouse retinal explants specifically abolished RGCs axonal growth induced by SFRP1, but had no effect on RGCs differentiation and it did not modify the effect of Shh or Netrin on axon growth. Taken together these results demonstrate that expression of Pax6 is necessary and sufficient to render postmitotic neurons competent to respond to SFRP1. These results reveal a novel and unexpected function of Pax6 in postmitotic neurons and situate Pax6 and SFRP1 as pair regulators of axonal connectivity.