Paper about extracellular ephys is published!
Our paper has been published in Nature Communications!
There are many cell types in the brain. Cells and in particular neurons could be classified by how they look like - morphology; how they respond to electric current - electrophysiology; and what genes do they read from the genome - transcriptomics. However, when we are observing cells in living conditions, so-called in vivo (in a living organism) - it becomes almost impossible to understand what cell type has been recorded. For example, if you install an electrode into the mouse visual cortex, you would see different electric impulses called spikes. The problem is the link between cell classes and extracellular spikes. What cells generate what impulses and where spatially in the brain?
In our work we made one step closer to understanding what extracellular signals one could measure with electrodes in vivo. To find a way to better interpret the signals we turned into computational modeling. First we developed computational models that describe the electrical activity of neurons. These models were developed for cells of different shapes and electric properties (cell classes) and used them to simulate spike generation. Using Physics first principles, electrophysiology and morphology of cells we simulated electric cell behavior of neurons and computed the extracellular electric fields these neurons generate. Then we compared these simulations with in vivo recordings from the mouse visual cortex, when the mouse is awake and is watching visual stimuli. What the mouse was watching is not that important, as we have concentrated on the extracellular signals.
Using neuron models and data analysis we found that we need to analyze the extracellular signals differently than it was analyzed before. We found that addition of certain features of extracellular spikes to our analysis could help to distinguish one cell class from another so we could better interpret the experimental recordings from the mouse. To validate these features we have used additional experiments, where we have stimulated particular cell classes using optogenetics. In other words we have validated our findings, where we stimulated particular neuron types in the mouse brain in vivo. These results showed the power of computational modeling, data analysis and experimental design to identify hidden features in the neural data.
The closer we look into the brain, the more we realize how complex is the regulation of neural activity. For a long time scientists were thinking that there are only 2 types of neural cells: excitatory and inhibitory. Now we know there are many more types (about 5000 in mouse brain), but what these types do in vivo and why there are so many remains a mystery. Using more nimble methods to identify cell types in a living organism moves us one step closer to understanding what the brain does on the level of individual cells. This understanding could help us build more efficient brain-computer interfaces and help to link neutral activity with cognitive functions.
