Human Brain Atlas Complete - Allen Institute for Brain Science

Jonah Lehrer's interview with Allan Jones, the CEO of Allen Institute for Brain Science:

"LEHRER: Let’s begin with the basics. How did you make this map?

JONES: We work with medical examiners and existing brain banks on both coasts. The brains we obtain through these sources have to meet strict criteria for a normal brain (typically accidental death or one caused by something other than brain injury or disease, such as cardiac arrest) and there has to be enough time after time of death to collect, process and freeze the brain within a 24 hour window. After consent is obtained from next of kin, we first get an MRI of the brain.  This serves as a “scaffolding” framework for us to ultimately hang the detailed gene expression data. The brain is then removed, and slabbed into 1 cm slabs and quickly frozen. The frozen slabs are shipped to the Institute, where we further process them: sectioning the large slabs and staining them with histological stains so that we can determine specific anatomic locations for sampling, then partitioning the slabs into more manageable sized 2” x 3” pieces, then taking these pieces and thinly slicing them to put on special microscope slides. The slides are then subjected to laser capture microdissection (LCM); technicians and expert anatomists work together to draw regions of interest on a computer screen connected to an LCM microscope, then a laser precisely cuts the delineated areas and these fall into a small plastic tube. RNA is extracted from this tissue, and a quantitative read-out from the RNA is obtained on a DNA micro-array that takes assays of the ~25,000 genes in the human genome. For each human brain we generated over 50 million gene expression data points across ~1,000 anatomic regions using this methodology.

LEHRER: You found that these two brains reveal a 94 percent similarity in average gene expression. Were you surprised by that number? How does it compare to the mouse map data? It strikes me as somewhat sobering that all of our individuality can be compressed into a mere 6 percent of variation in gene expression. Finally, have you begun analyzing these differences?

JONES: Given that we are looking at a small number of brains, we chose to focus on the similarities rather than differences at this point. We are encouraged by the finding (basically, the number is the average across all structures of the percent of genes that are expressed in common for each structure across the first two brains).  Indeed, it will be fascinating to start to dig into the areas of similarity and difference, both in structure and in functional class of gene. We are just starting to delve into this level of detail for analysis and will likely have more to say on it in the future.

LEHRER: Frankly, I was astonished that 82 percent of all human genes are expressed somewhere in the brain. This seems to me like yet another reminder that the brain is one awesomely complicated piece of meat.

JONES: This number is almost exactly what we found in the mouse, so that is good corroboration. When you think about the complexity of the functions of the brain, and the variety of different cell types found within the brain (there are blood cells, epithelial cells lining the blood vessels, pockets of dividing adult stem cells, plus all of the flavors of neurons and glia) it’s not quite as surprising to see how much of the genome is used to serve the brain."