With the recent advances in deep learning enabling rapid and accurate identification of complex structures, it’s application to in vitro patch-clamp electrophysiology was inevitable. We are exploring the application of these techniques to enhance the capabilities of our fully-automatic patch clamp robots. For example, we have demonstrated fully automatic, real-time detection of healthy neurons within traditional DIC images and vision-based techniques to correct for hardware error stack-up during pipette localization.
In vivo patch-clamp is the gold standard for intracellular recordings, but it is a very manual and highly skilled technique. We have created the most automated in vivo patch-clamp robot to date, by enabling production of multiple, serial intracellular recordings without human intervention. Our robot automates pipette filling, Ag/AgCl wire threading, pipette positioning, neuron hunting, break-in, delivering sensory stimulus, and recording quality control, enabling in vivo cell-type characterization.
Intracellular patch-clamp electrophysiology, one of the most ubiquitous, high-fidelity techniques in biophysics, remains laborious and low-throughput. While previous efforts have succeeded at automating some steps of the technique, we have created a robotic ‘PatcherBot’ system that can perform many patch-clamp recordings sequentially, fully unattended. Comprehensive automation is accomplished by outfitting the robot with machine vision, and cleaning pipettes instead of manually exchanging them. The PatcherBot can obtain data at a rate of 16 cells per hour and work with no human intervention for up to 3 h. We have demonstrated the broad applicability and scalability of this system by performing hundreds of recordings in tissue culture cells and mouse brain slices with no human supervision. The system is potentially transformative for applications that depend on many high-quality measurements of single cells, such as drug screening, protein functional characterization, and multimodal cell type investigations.
Serial section electron microscopy (ssEM), a technique where volumes of tissue can be anatomically reconstructed by imaging consecutive tissue slices, has proven to be a powerful tool for the investigation of brain anatomy. Between the process of cutting the slices—or “sections”—and imaging them, however, handling 100-106 delicate sections remains a bottleneck in ssEM, especially for batches in the “mesoscale” regime, i.e.,102-103 sections. Our lab is developing a tissue section handling device that transports and positions sections accurately and repeatably for automated, robotic section pick-up and placement onto an imaging substrate.
Recent advancements utilizing induced pluripotent stem cell-derived (iPSC) epithelia have made disease modeling and cell therapy for many, previously untreatable diseases possible. However, current electrophysiology techniques – used to validate epithelial function – are painstaking and require a highly trained user; limiting experimental throughput. We are developing new techniques and devices that enable both high-throughput and high-quality electrophysiological measurements. Our lab has built a robot that automates intracellular electrophysiology of epithelia that automates the pipette insertion process and, simultaneously improves throughput. We are also exploring new electrochemical impedance spectroscopy (EIS)-based techniques to extract high resolution, membrane-specific properties non invasively. These techniques could be used as the basis for at-line functional characterization of epithelia in all future iPSC-based therapies.
Brain organoids have allowed neuroscientists to make valid predictions about human neurodevelopmental diseases on the basis of organoid morphology, cellular distribution and composition, and gene expression. Following this trajectory, neuroscientists have proposed brain organoids as a model of human synaptic function. However, this approach is hindered by a primary limitation: techniques to characterize the electrophysiology of living synapses are far too slow and laborious to be applied to comparative or longitudinal studies required to validate organoid models and generate new hypotheses. We are currently addressing this limitation to the growth of brain organoids as a model of synaptic physiology by advancing the throughput and quality of automated multiple patch clamping in intact brain organoids.
We are extending on our prior pioneering work on automated patch clamping robots by developing a robotic patch clamp system for optogenetic tool screening. We will then apply this technology to improve the kinetics of different optogenetic molecules, and also seek to improve the performance of an important class of optogenetic tool – improving the selectivity and conductances of light-gated potassium channels.
Our lab, in conjunction with our collaborators at Emory (Andrew Jenkins and Steve Traynelis) have added improvements to the patcherBot system in order to improve neuropharmacology screening. These improvements include manipulation of heterologous cells and control of submillisecond solution exchange to mimic the speed at which neurotransmitters are released and removed from the synaptic cleft, where they activate ligand-gated ionotropic receptors. The “pharmaBot” system can perform typical ligand-gated ionotropic receptors experimentation protocols autonomously that allows for a high experiment completion success rates and can reduce the operator’s effort substantially.
Our lab, in collaboration with GTRI and other labs at Georgia Tech (David Hu Lab, Eric Vogel Lab, food and Processing) is developing the next generation sensor for Weapons of Mass Destruction (WMD). This deployable device will collect air particulate and provide crucial WMD agent identification, location and concentration information vital to battlefield decision making. We are bridging our microfluidics expertise for the filtration aspect of the project and look for the projects’ applications in other areas such as point-of-care diagnostics, pollution control, among others.