LABEL-FREE, HIGH-SPEED QUANTITATIVE IMAGING OF ASTROCYTE-NEURON NETWORKS WITH OPTICAL DIFFRACTION TOMOGRAPHY AND MACHINE LEARNING
Abstract
We propose to apply a combination of ultrasensitive optical detection methods and machine learning to establish a label-free platform for visualizing and characterizing activity of neuronastrocyte networks with single-cell resolution. Research on glial cells over the past three decades has confirmed the critical role of astrocytes in brain physiology. The concept of the tripartite synapse reflects the important role of astrocytes in exchanging information with the synaptic neuronal elements, responding to synaptic activity and regulating synaptic transmission. Yet, our knowledge of neuron-astrocyte coupling in complex networks remains limited and fragmentary. Despite the advent of new high-resolution microscopic methods, many approaches require high and potentially damaging excitation intensities and/or contrast agents that may perturb the native physiological activity. In this work, we will leverage optical diffraction tomography (ODT) that offers nanoscale sensitivity to morphology and dynamics, non-destructive imaging of transparent biological structures and quantitative signals based on refractive index (RI) contrast. By using deep convolutional neural networks that allow automatic recognition of key biological traits encoded in RI tomograms, we seek to shed new light on the following fundamental questions. First, we aim to examine morphological heterogeneity in astrocyte populations and visualize and quantify the lipid content in a label-free manner, the latter having a direct bearing on the protective role of astrocytes in the central nervous system. In the second aim, we will focus our ODT-deep learningefforts on characterization of neural networks with the spatial resolution of a single synapse and the scale to image large ensembles of synapses. Building on the results of these two aims, we finally seek to extend our platform to examining neuron-astrocyte co-cultures with a special focus on understanding the metabolic coordination between them. By measuring fatty acid transfer using correlative ODT and fluorescence microscopy, we aim to uncover neuron-astrocyte coupling of lipid metabolism.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 20, 2023
- Source ID
- FA95502210334
Entities
People
- Ishan Barman
Organizations
- Air Force Office of Scientific Research
- Johns Hopkins University
- United States Air Force