top of page



One of the most fundamental decisions axons make is whether or not to cross the border between the central and peripheral nervous system (CNS and PNS, respectively). While the vast majority of axons in the vertebrate nervous system do not traverse the CNS-PNS boundary, motor neurons in the spinal cord and hindbrain project axons into the periphery. Conversely, sensory neurons in the trigeminal ganglion and the dorsal root ganglia (DRGs) send axon branches into the brain and spinal cord. The mechanisms that prevent or allow axon growth across the CNS-PNS boundary remain largely unexplored.


The meninges envelop the entire CNS and are therefore in an ideal position to regulate axon behavior at the CNS-PNS interface. We found that the meninges surrounding the developing neural tube secrete diffusible guidance cues for several classes of spinal cord neurons. Importantly, these guidance activities are consistent with the respective axonal trajectories relative to the meninges in the developing nervous system. Our findings support the idea that meningeal guidance cues contribute to nervous system wiring, and we are investigating the molecular mechanisms through which the meninges sculpt neuronal connections and regulate axon behavior at the CNS-PNS boundary.


In a second line of investigation, we identified the motor neuron cell surface protein TAG-1 as a key multifunctional regulator of neuronal migration and axon guidance at the CNS-PNS border. In mice lacking TAG-1, motor neuron cell bodies aberrantly leave the CNS, and motor axons are misrouted immediately after spinal cord exit. We are studying the cellular and molecular mechanisms through which TAG-1 coordinates multiple steps in neuronal wiring.

Ectopic motor neurons


NELL2 repulsion



Slit proteins are repulsive axon guidance cues acting through the receptors Robo1 and Robo2. Slit-Robo signaling regulates axon guidance in a variety of neuronal populations and can also induce axon branching, fasciculation, and cell migration. In mammals, the divergent Robo family member Robo3 antagonizes Robo1/2-mediated Slit repulsion and potentiates attraction to Netrin-1 through the Netrin receptor DCC.


We found both in vitro and in vivo evidence for the existence of a novel Slit receptor. We are investigating the identity of this receptor, and we aim to understand which aspects of Slit function are mediated by this molecule.


We also discovered a novel repulsive guidance molecule, NELL2, that acts through the previously orphan Robo3 receptor and helps guide axons to the spinal cord midline. Our results indicate that Robo3 is an integrative hub for three different axon guidance signaling pathways. We seek to understand how Robo3 can mediate its multiple distinct functions (silencing Slit repulsion, potentiating Netrin attraction, and mediating NELL2 repulsion) by studying the signal transduction events downstream of Robo3. We are also characterizing the contribution of NELL2 and Robo3 to different aspects of nervous system wiring. Lastly, we aim to gain a deeper understanding of Robo evolutionary history and the importance of Robo family member specialization in mammals.



The perception of touch, pain, and other somatosensory modalities relies on information provided by receptors in the body periphery. However, our understanding of somatosensation is still incomplete, partly because the identities and wiring patterns of spinal cord neurons receiving information from modality-dedicated peripheral neurons have not been fully characterized.


Dorsal commissural neurons (C-neurons) project axons across the ventral midline of the spinal cord and relay sensory input from DRG neurons to the brain stem, cerebellum, and thalamus, but also make local connections within the spinal cord. C-neurons are a heterogenous population of cells with respect to their cell body position, axonal trajectory on the contralateral side of the spinal cord, and neurotransmitter phenotype, and a detailed understanding of their connectivity, especially their presynaptic inputs, is lacking.


We are systematically mapping the input-output relationship of distinct C-neuron subpopulations using a combination of mouse genetics,  viral tracing, and in toto imaging. We aim to identify the classes of sensory neurons with synaptic inputs to C-neurons and understand how different somatosensory modalities are relayed by subgroups of C-neurons with projections to distinct targets. We will combine our circuit mapping results with functional studies to gain insights into mechanisms of somatosensory information processing. We also employ single-cell RNA sequencing  to identify the transcriptional programs that specify the wiring patterns and functions of different C-neuron subtypes.

Commissural neurons

Denervated NMJ.jpg

NMJ denervation



Amyotrophic lateral sclerosis (ALS) is a devastating neurodegenerative disease characterized by progressive degradation of neuromuscular connectivity, leading to widespread paralysis and ultimately death of affected individuals. Effective therapeutic intervention in ALS presents an urgent, unmet medical need, and understanding disease mechanisms is therefore of paramount importance.


We are using knock-in models in mice to study ALS etiology and identify potential therapeutic targets. We found that mice carrying disease-causing mutations identified in human patients display hallmarks of ALS pathology, including motor neuron death and progressive denervation of neuromuscular junctions (NMJs). These findings establish our knock-in mice as accurate disease models, and we are continuing to study the full spectrum of ALS-like defects in these mouse lines. Further, we are using our mouse models as a platform to test various genetic manipulations for their ability to ameliorate ALS symptomatic progression.

bottom of page