The focus of our research is to address:
We take a multidisciplinary approach that include genetics, cell biology, developmental biology, biochemistry, advanced imaging (for neuronal structures and activity), electrophysiology, computation (including machine learning and computer vision) and behavioral studies.
Dendrites and axons are the input and output compartments of a neuron, respectively. The distinct morphology of a neuron’s dendrites and axons is fundamental to the assembly of neural circuits. Using Drosophila systems, we identified molecular and cellular mechanisms that regulate dendrite and axon development differently. Our lab also contributed to the understanding of the function, regulation and diversity of the Golgi apparatus in neurons and developed molecular tools for studying the Golgi complex in vivo in multi-cellular organisms.
Learn More:
Building on our strength in cell and developmental biology of neurons as well as Drosophila genetics, we studied the molecular, cell biological and developmental mechanisms that assembles neurons into functional neural circuits. Our lab reported the first neural-activity-dependent topographic map in Drosophila (Yang et al., 2013; Kaneko & Ye, 2015) and used this system to investigate the molecular mechanism underlying the assembly of neural circuits. We also discovered a novel form of sensory-pathway-specific plasticity during the development of Drosophila larvae (Kaneko et al., 2017).
Learn More:
Currently, we are trying to discover the principles that govern the assembly of neural networks beyond those that regulate neuronal development (e.g., dendrite and axon development) and the connections between two groups of neurons (e.g., axon targeting and synapse development). This is a new frontier in developmental neuroscience and requires multidisciplinary approaches, including molecular, cellular, genetic, developmental, physiological, behavioral and computational approaches, which we have established in our lab.
Learn More:
We discovered that the expression level of Down syndrome cell adhesion molecule (Dscam) instructs the growth of axon terminals in Drosophila (Kim et al., 2013). The gene encoding the human homolog of Dscam (called DSCAM) is on chromosome 21, which is present as three copies in Down syndrome individuals. DSCAM levels in the brain are changed in Down syndrome and several other brain disorders. We further found that the cytoplasmic tyrosine kinase, Abl, is required for dysregulated Dscam to promote presynaptic terminal growth and that Abl inhibitors attenuates the neurological defects caused by Dscam overexpression (Sterne et al., 2015). Recently, we demonstrated that DSCAM overexpression leads to excessive GABAergic innervation and synaptic transmission in the neocortex of mouse models of Down syndrome (Liu et al., 2023). Currently, we are examining other genes on human chromosome 21 for potential effects on brain development. This line of research uses Drosophila and mouse systems to complement each other.
Learn more:
Our research on neural circuits has taken us to study the mechanisms that underlie neural computation. We recently discovered a neural basis for categorizing sensory stimuli to enhance decision accuracy (Hu et al., 2020). In this work, we established a method for analyzing neural activities in the entire CNS of the Drosophila larva and developed novel computational tools to identify CNS regions involved in sensory processing. We also established patch-clamping recordings of CNS neurons in the larval somatosensory system and discovered low-threshold mechanosensory neurons inhibit the nociceptive pathway to filter out weak nociceptive inputs (Pan et al., 2023). In addition to studying neural computation in the Drosophila somatosensory system, we are exploring the possibility of applying the neural algorithms to robotics for testing these algorithms and for improving robot performance.
Learn more:
A major goal of neuroscience is to understand the neural basis for behavior, which requires accurate and efficient quantifications of behavior. To this end, we recently developed a software tool for automatic identification and quantification of user-defined behavior through artificial intelligence (Hu et al., 2023). This tool is not restricted to a specific species or a set of behaviors. The updated version (LabGym2) can analyze social behavior and behavior in dynamic backgrounds. We are further developing LabGym and other computational tools for behavioral analyses in wild animals and in medicine.
Learn more:
