Background We are interested in the development and function of neuronal circuits. Our approach consists of generating genetically modified mouse lines, in which specific genes are ablated and replaced by histochemical reporters that allow the visualization of the individual cells in which the gene manipulation has been carried out. This manipulation allows us to describe the anatomy, and developmental history of the manipulated neuronal populations, in the context of either wild type or mutant gene dosages, and thereby gain insights both about the function of the cell within the circuit and the gene within the cell. Specifically, we had previously generated conditional knock-in reporter lines for a family of three transcription factors, Brn3a, Brn3b and Brn3c, which are expressed in Retinal Ganglion Cells (RGCs). We now describe the use of these lines towards the definition of RGC cell types, their development, and their role in neuronal circuits. Results a) Previously we had described the specific RGC types expressing the transcription factors Brn3a, Brn3b and Brn3c. Using our conditional reporter alleles we also have described a similar combinatorial role of Brn3s in other sensory projection neurons of the somatosensory and auditory system. Thus, Brn3cAP Dorsal Root Ganglia (DRG) neurons belong largely to the peptidergic nociceptor family, Brn3bAP DRG neurons fall in the mechanoreceptor class, and Brn3aAP DRGs are covering most decribed DRG cell types, including mechano, proprio and nociceptors. Whereas no significant defects were found in Brn3bAP/KO and Brn3cAP/KO DRGs from the anatomical perspective, Brn3aAP/KO DRGs exhibited a specific loss of hair follicle associated sensory fibers, coupled with a depletion of projections to the dorsal horn of the spinal chord, most likely signifying a dramatic loss of mechanoreceptor DRGs. Together with our studies in RGCs, these findings support a model in which Brn3s participate in a combinatorial code of transcriptional regulation of sensory projection neurons, and help establish their cell type identity and regulate the gene expression of molecules required for the correct axonal targeting, dendritic arbor formation, and in many cases survival of these neurons. This work, initiated in Johns Hopkins, was finalized at the NIH, with the assistance of Melody Shi and Oluwaseiy Motajo. (Badea et al. 2012, J. Neuroscience) b) It remains unclear how the partial overlap of expression and function of Brn3s contributes to RGC diversity. A host of previous findings point at some degree of genetic interaction between the three Brn3 factors during RGC specification. It is however unclear how Brn3b might regulate Brn3aAP or Brn3cAP RGCs. Moreover, genetic interactions between pairs of Brn3 transcription factors have not been carefully characterized. We therefore have generated Brn3bKO/KO;Brn3aKO/AP, Brn3cKO/KO;Brn3aKO/AP and Brn3bKO/KO;Brn3cKO/AP double knock-outs, in which RGCs were labeled respectively by Brn3aAP, Brn3aAP or Brn3cAP, and studied the effects of deleting each Brn3 alone or in combination on the selected Brn3AP expressing population. The first dramatic finding is that many Brn3aAP RGCs survive in the absence of Brn3b, however combined loss of Brn3a and Brn3b results in an almost complete depletion of Brn3aAP RGCs, and RGCs in general. This finding strongly argues for distinct roles of Brn3a and Brn3b in RGC development. We also find a a reduction in Brn3cAP RGCs, in Brn3bKO retinas, coupled with a distinct and very specific deletion of the OFF morphology of Brn3cAP RGCs, and an overall increase in dendritic arbor area increase in surviving Brn3cAP RGCs. This is consistent with the fact that the OFF Brn3cAP RGC cell type is the only one also positive for Brn3b. No genetic interactions were observed between Brn3b and Brn3c or Brn3a and Brn3c with respect to RGC development, as double knock-outs of the two pairs of transcription factors do not exhibit distinct phenotypes when compared respectively to the Brn3b or Brn3a single knock-outs. Part of this study, carried out by M. Shi, S.R. Kumar, O. Motajo and T. Badea, has been presented as an abstract at ARVO 2011. c) To understand the developmental history of our cells, we are using sparse labeling strategy to label early (Embryonic 12-15) stages in RGC development. We can thus study the trajectories of individual Brn3AP RGC axons as they progress through the various decision points inside and outside the brain (intraretinal, optic disc, optic stalk, optic chiasm and tract, diencephalon, pretectal area, optic tectum), under wild type and mutant circumstances. In addition, in collaboration with Carol Masons group, we are analyzing the presence of ipsi- versus contra-laterally projecting axons at the earliest stages of optic chiasm formation. In addition, we have observed expression of Brn3b in many of the central targets of sensory projection neurons: - Superior Colliculus for RGCs, - nucleus of cranial nerve V for the Trigeminal Ganglion, - auditory and vestibular nuclei for the Spiral and Scarpa ganglia, and dorsal horn of the spinal chord for the DRG. These findings further strengthen the link between Brn3 transcription factors and the major sensory pathways, suggesting a marking of the ascending sensory pathway by Brn3s. We are now in the process of validating the identity of the various nuclei with molecular markers, and describing the single cell anatomy of the individual neurons in these nuclei. d) To understand the molecular requirements for RGC cell type determination, we are conducting a screen for genes specific for RGCs expressing Brn3a, Brn3b or Brn3c at ages relevant for neuronal arbor formation (E15 for axon guidance and P3 for dendrite formation and synapse formation). This is done by magnetic separation of Brn3AP RGCs, using monoclonal antibodies against AP, a GPI linked surface molecule, followed by RNA extraction and RNA sequencing in the N-NRL genomics core. Data sets for Brn3a and Brn3b RGCs at P3 and Brn3b at E15 have already been generated, and an impressive set of candidate genes with functions relevant to cell morphology development identified. Validation will be performed on a selected subset of these candidates, and will involve in situ hybridization in cryosections of retinas of Brn3AP mice at P3. e) We are expanding our genetic toolbox by generating mouse lines dependent on Dre recombination. Dre recombinase is an equally effective homologue of Cre recombinase, but operating on distinct molecular target sites. We have generated two Dre dependent conditional knock-in Cre expressing mouse lines. In these lines, named Brn3aCKOCre and Brn3bCKOCre, the Brn3a or Brn3b genes are flanked by Dre recombinase target sites (roxP sites), and followed by the Cre recombinase. When exposed to Dre recombinase activity, the endogenous genes are removed, and replaced by the Cre recombinase, resulting in conditionally knocked-in Cre alleles. The Cre, now expressed specifically from either the Brn3a or Brn3b locus will activate any desired downstream target. Using these lines, we will be able to specifically remove Brn3a or Brn3c from RGCs that also express Brn3b using the Brn3bCKOCre in conjunction with the Brn3aCKOAP or Brn3cCKOAP alleles. At this point, we have generated and confirmed by southern blotting the Brn3aCKOCre and Brn3bCKOCre lines. Significance This past year we have made significant advances in understanding the mechanisms by which Brn3 transcription factors regulate RGC and indeed sensory projection neuron development. In addition, downstream transcriptional targets of Brn3, which are potentially involved in the development of RGC types.
Showing the most recent 10 out of 25 publications
