Neuroscience textbooks present regions of the brain as relatively homogenous units composed of a few different discrete types of cells each: the cortex has pyramidal cells, the cerebellum has Purkinje cells, the dentate gyrus has granule cells, to name a few, and all are sparsely dotted with interneurons. In reality, most regions of the brain are much messier, containing a tangled mix of dozens of cell types varying along numerous spectra, such as morphology and gene expression. If traits we observe in neurons are continuous instead of discrete, and these neurons must be organized on many continua, two obvious questions arise: How do we define a cell type, and are current classifications adequate to describe the full range of neurons we can find in the nervous system?
Dr. Hongkui Zeng, the Executive Director of Structured Science at the Allen Institute for Brain Science, has devoted her career to answering these questions. She is primarily interested in how a neuron’s genes determine its physiology and connections with other types of neurons, and she has created many public datasets in the pursuit of these answers. In the process, she has also developed many high-throughput pipelines and tools relied up on by neuroscientists all over the world, and she has led numerous large-scale projects including the Human Cortex Gene Survey, Allen Mouse Brain Connectivity Atlas, and the Mouse Cell Types and Connectivity Program.
At the beginning of 2016, she took an important step towards answering these questions in her paper “Adult mouse cortical cell taxonomy revealed by single cell transcriptomics,” published in Nature Neuroscience. In this paper, her team at the Allen Institute set out to characterize all of the cell types in the primary visual cortex, one of the most well-defined and study regions of the brain. To do this, she ran single-cell RNA sequencing on many genetically-modified mice where cells expressing certain genes would glow, allowing her team to extract these glowing cells in very small amounts and find underlying patterns of gene expression. By identifying patterns exclusive to certain cells, they were able to characterize 49 cell types within the primary visual cortex. Many of these cell types have not previously been described, here or anywhere else in the brain, and in most cases the genes defining these groups had not been used as a molecular marker previously. Interestingly, these cell types were not discrete, but instead in two tiers: a core tier central to each group, and an intermediate tier displaying gene expression partially characteristic of multiple cell types.
Figure 1: Initial determination of cell types in primary visual cortex.
Using these new markers for these new groups, her team used in-situ hybridization to locate these cell types within the cortex. This allowed them to place these cell types in the context of the cortex, which helps specify their function and their connectivity based on their home layer. These groups were also linked by their respective intermediate cells similar to both groups, which allowed lineage tracing and linkage to be performed, assembling these core groups into larger meta-groups based on close clustering of these core groups.
Figure 2: Locations and relationships of cell types in primary visual cortex.
Finally, she and her team injected a viral retrograde tracer into the contralateral primary visual cortex and ipsilateral thalamus, two regions that primary visual cortex neurons project to. Cells would project to the contralateral visual cortex or the ipsilateral thalamus, but not to both. This allowed them to run RNA sequencing on the groups of neurons projecting to each region and then classify each group by what region it projects to, adding another dimension to these groupings.
Dr. Zeng’s findings provide a great new resource and framework for other investigators. In this study, she was able to identify cell types in the primary visual cortex with the greatest granularity recorded thus far. She was then able to characterize all these groups by differential gene expression, degree of relatedness, their location in the cortex, and the location of their projections. These findings lay the groundwork for similar studies in other regions of the brain, and could hopefully one day provide insight into what makes exactly gives these diverse groups of neurons their unique properties.
James Howe is a first-year neuroscience Ph.D. student currently rotating in Dr. Rusty Gage’s laboratory.