Current Projects

We use the mouse to define how the mammalian nervous system develops. Through chemical mutagenesis, we have many recessive mutant lines with specific defects in the nervous system, and we are using deep resequencing to identify the genetic lesions in the lines. This approach is especially fruitful since it permits us to focus on novel genes.

For example, one of the mutant lines affects a small GTPase, Arl13b, which is localized to cilia and disrupts Sonic Hedgehog (Shh) signaling, a critical pathway in spinal cord development. In humans, defects in Arl13b can cause Joubert Syndrome and may underlie other cilia-related diseases. We combine biochemical, molecular, cellular, and genetic approaches to figure out the mechanism underlying Arl13b function, as well as the other genes identified in the screen.

Figure 1: ENU Mutagenesis

Figure 1. ENU mutagenesis screen. Schematic depicting 3 generation cross to identify autosomal recessive mutations that disrupt neural development.

Neural Development

The nervous system is patterned along the anterior-posterior axis and the dorsal-ventral axis. While some of the pathways that specify the various cell fates in the developing nervous system are known, our current knowledge has clear gaps. We don't understand the mechanism by which all the proteins function in specific steps of the known pathways. We also know that other signaling pathways are involved, but we do not yet know their identity.

In order to fill these gaps, we use phenotype driven screens in the mouse and have identified 14 recessive mutations with morphological defects in neural development midway through gestation. Using deep resequencing, we have found the putative causal mutation in several lines. Interestingly, in several cases, the affected genes are also expressed in the adult cerebellum, so we are working out their function in cerebellar development.

Figure 2: Mutant Lines and Expression in the Cerebellum

 Figure 2: Embryonic day 10.5 (A)Wild type and (B-E) mutant embryos for four of the newly identified lines showing abnormal neural development. (F-I) In situ hybridization of postnatal day 55 mouse brain from the Allen Brain Atlas showing the expressive pattern of the affected gene in each of the mutant lines.

Dorsal-Ventral Patterning of the Spinal Cord

The various cell types within the spinal cord are specified based on their position along the dorsal-ventral axis. Sonic Hedgehog signaling is critical for specifying ventral cell fates. Indeed, a major focus of our lab is a mutant line in a small GTPase called Arl13b. Loss of Arl13b results in low-level constitutive Shh signaling which means that too many motor neurons (MNs) are specified and that there is no floor place. Cilia are required for Shh signaling, and we discovered that Arl13b is highly enriched in cilia. Normally, the components of Shh signaling move dynamically in and out of cilia in response to Shh stimulation. However, the dynamic movement does not occur in the absence of Arl13b. We are working out the details of the mechanism through which Arl13b regulates Shh signaling.

Figure 3: Dorsal Ventral Patterning of the Neural Tube

Figure 3. Signals from the roof plate (purple) and floor plate (red) induce the expression of transcription factors that specify cell fate. Transcription factor domains are indicated in colored boxes; cell domain names are described by colored circles.

Figure 4: Hnn Patterning

Figure 4. Disrupted neural patterning in Arl13bhnn mutants. Immunofluorescence of cell markers in wild type (top) and Arl13bhnn (bottom) embryos. Arl13bhnn mutants display an expansion of the pMN (Olig2+) and motor neuron cells (Isl1/2+ and HB9+) in the absence of a floor plate (Shh+).

In contrast to ventral neural tube patterning, the signals that specify dorsal cell fates are less well understood. There are 8 known types of dorsal interneurons generated from distinct progenitor domains along the dorsal-ventral axis of the dorsal neural tube. Both Wnt and Bmp signaling play important roles in specifying dorsal cell fates, but the details are less clear compared to ventral neural patterning. To address this, we incorporated a Pax3-GFP transgene into our screen (courtesy of Dr. Paul Trainor, Stowers Institute for Medical Research). This transgene is normally expressed exclusively in the dorsal spinal cord, so we looked for disruption of its expression in embryos that genotype positive for the transgene. We found three novel lines and are in the midst of finding the casual mutation through deep resequencing.

Figure 5: Pax3-GFP Expression in Mutants

Figure 5. Heritable lines displaying aberrant Pax3-GFP expression. These are the lines that display either abnormal (M2 and AB4) or a lack (AM3) of Pax3-GFP expression in the neural tube at embryonic day 10.5.