Synaptic mechanisms of temporal coding in the Drosophila mushroom body
1 Yale University
Kenyon cells (KCs) in the Drosophila mushroom body encode odor identity as part of the circuit responsible for olfactory associative learning. Each of ~2,000 KCs receive sparse and divergent input from a pseudo-random set of ~6 out of the 52 types of antennal lobe projection neuron (PN), and only 5-10% respond to an odor stimulus. Most prior work has treated KC responses as a binary on or off switch, showing that this circuit architecture implements a pattern separation algorithm that optimally facilitates learning at the convergent KC output. However, KCs exhibit temporally complex and diverse odor responses that could function to further enhance the encoding and separation of odor identity. Here, we are investigating the biophysical mechanisms contributing to this temporal diversity with a focus on short-term plasticity (STP) at the PN-KC synapse. After establishing a whole cell recording from a KC, we use 2-photon optogenetic stimulation of the PNs to identify and characterize each synapse. Measuring the induced synaptic currents indicates the strength of synaptic facilitation or depression. We are additionally using Bayesian inference to fit our data to the Tsodyks-Markram model of STP and create a network model where we can ask how altering STP properties causally changes spiking behavior. Our results are showing a high degree of variability in STP across synapses, including ones from the same PN onto different KCs. We are hypothesizing that each PN-KC synapse is randomly assigned a facilitating or depressing phenotype, which generates unique temporal integration windows across the population to enable temporally diverse odor responses.