Cell Type Look-up Table

The cell type look-up table dataset is a collection of recordings taken from the mouse striatum with the goal of collecting a ground truth data set detailing the responses of different striatal cell types. Data was collected using extracellular electrophysiology. We used transgenic tools to target expression of light-gated ion channels (e.g. CoChR, ChrimsonR, and ChRmine) in specific cell types. During recordings, mice were awake and head-fixed on a running disk.

Background

It is a well known fact that neurons in the brain are not homogeneous: they differ in terms of morphology, connectivity, gene expression, and other factors. These differences imply a difference in function, and indeed, different “types” of neurons have been found to play distinct roles in the computations performed by the brain, whether that be sensory processing, motor output, or decision-making [Cox and Witten, 2019]. As such, any modern study on the function of the brain needs to take cell type into account.

However, determining the identity of individual neurons can pose a challenge in the middle of a recording. This is especially true for extracellular electrophysiology, as we have no access to the information usually used to determine a cell’s type, such as morphology or gene expression. This is a big blow to a technique that otherwise has unparalleled reach in terms of temporal resolution and access to different brain areas.

But all is not lost! After all, differences in a cell’s morphology and gene expression should also lead to differences in physiological properties. That is the purpose of this data set: to collect electrophysiological data from units belonging to known cell types (using a technique known as “optotagging”) and determine what, if any, differences exist between members of different cell types.

Data set

The current data set you have access to contains optotagging experiments collected from the striatum. The striatum is part of the basal ganglia and is associated with many functions, including locomotion, decision-making, and reward. Of particular interest is the fact that medium spiny neurons, which make up 95% of the neurons in the striatum, can be cleanly divided into two cell types based on their projection and gene expression patterns. These are the “direct pathway” neurons, so called because they promote movement, and the “indirect pathway” neurons, so called because they inhibit movement. The direct pathway neurons express the dopamine D1 receptor and project to the internal globus pallidus and to substantia nigra, and are often referred to as ‘D1 cells’. The indirect pathway neurons express the dopamine D2 receptor and project to the external globus pallidus, and are often referred to as ‘D2 cells’.

These two cell types have been heavily studied by basal ganglia physiologists, and are known to play different roles in behavior, and as such are a fascinating topic of study. But there’s a major issue with them: because they are morphologically so similar to each other, they are (currently) impossible to distinguish in electrophysiological recordings without the use of optotagging! Thus, this data set hopes to collect a ground truth of tagged D1 and D2 cells to determine if there is any physiological difference between these two cell types that would aid in telling them apart in unlabeled recordings.