The Allen Brain Atlas just launched their first set of gene expression maps in the adult human brain, based on microarray data from over 700 different anatomical locations. It promises to be an invaluable resource for scientists trying to figure out how a text of base pairs constructs the most complicated machine in the known universe. I wrote about the construction of the human brain atlas last year in Wired, if you'd like to learn more about how the map was made. Although these genetic maps are just a first draft - one researcher at the Allen Institute compared them to those 15th century sketches of the New World - the data has already taught us plenty of interesting things about the three pounds of flesh in our head.
One unexpected--even disheartening--aspect of the Allen Institute's effort is that although its scientists have barely begun their work, early data sets have already demonstrated that the flesh in our head is far more complicated than anyone previously imagined.
The brain might look homogenous to the naked eye, but it's actually filled with an array of cell types, each of which expresses a distinct set of genes depending on its precise location. Consider the neocortex, the so-called CPU of the brain: Scientists assumed for decades that most cortical circuits were essentially the same--the brain was supposed to rely on a standard set of microchips, like a typical supercomputer. But the atlas has revealed a startling genetic diversity; different slabs of cortex are defined by entirely different sets of genes. The supercomputer analogy needs to be permanently retired.
Or look at the hippocampus, the crescent-shaped center of long-term memory. Until recently, this small fold of tissue in the middle of the brain was depicted as neatly divided into four distinct areas. But data from the atlas has rendered the old maps not only obsolete but flat-out misleading. Even a single hippocampal area can actually be subdivided into at least nine discrete regions, each with its own genetic makeup.
Scientists at the institute are just starting to grapple with the seemingly infinite regress of the brain, in which every new level of detail reveals yet another level. "You can't help but be intimidated by the complexity of it all," Jones says. "Just when you think you're getting a handle on it, you realize that you haven't even scratched the surface." This is the bleak part of working at the Allen Institute: What you mostly discover is that the mind remains an immense mystery. We don't even know what we don't know.
See the whole gallery of gorgeously gruesome photos over at Wired.
- Log in to post comments
Pretty cool, though things are even more complicated by the fact that those expression patterns probably change quite a bit with changes in brain state.
I wonder if this phenomenon occurs in the brains of other mammals.
@Jerry: The Allen Institute has already done a lot of the same work on the mouse brain ( http://mouse.brain-map.org/welcome.do ). Put in your favorite gene (such as crh) and it will show you expression patterns throughout the brain.
I thoroughly enjoyed your Wired article --- I hadn't realized how far this kind of scanning had already come. In the article you say that parts of the scanning are done at the cellular level. Am I understanding aright that the scientists at the Allen Institute are figuring out gene expression at the level of single neurons? Or is it for larger regions of the brain?
@Michael: You can resolve gene expression at the single cell level quite easily (this type of work has been done for years). The useful information for most researchers, however, is what do expression patterns look like in different brain regions. There are many reasons to want to know this, but one useful example is for predicting effects and side effects of different drugs (for example, if you have a drug that works on a gene product that's expressed in a brain region that's known to control a function such as movement (let's say you are developing a drug to help Parkinsonism), but it's also expressed in a brain region known to regulate memory, you might expect for there to be potential effects on memory processes).
This is still a pretty limited answer, but hopefully it will shed some light on your question.
@CMO: thanks for the response, which is illuminating. What I'm wondering is whether the Allen map in particular is at the single cell level? I'm curious in part because of work by other labs (e.g., the Lichtman lab at Harvard) on the human connectome. I'm interested in understanding how the Allen map relates to work on the connectome. If the Allen map is at the single-neuron level, then presumably that's a helpful start, although obviously it doesn't give you the information you'd like about axons and dendrites.
@Michael: With the Allen atlas, you are in fact able to identify single neurons (the nuclei are stained), but you don't haven any other information about that particular neuron. To answer questions of connectivity, the technique used by the Allen Inst. (in situ hybridization; ISH) is often coupled with other techniques. An example would be to inject a retrograde tracer into a brain region of interest (Lichtman's 2008 Nat Rev Neurosci review gives some explanation) and then you could use ISH to label neurons of interest (for example neurons of the hypothalamus that express crh) and see if they also are labeled with the retrograde tracer. If so, it would indicate that the hypothalamic neurons expressing crh send projections to the target brain region you injected.
As for the connectome work by Lichtman and others - they are using much more sophisticated techniques. Not to discount the work of the Allen Inst, but those maps aren't really suited for doing this kind of work, although they may provide preliminary data to help design the more complex type of work.
Again, because the atlas doesn't give you any other information about the cells labeled (aside from what brain region they are located in), it is fairly limiting in that regard. It is quite a useful resource for other reasons though.
@Michael: It also appears that for the human brain, they will using microarrays for data collection. In this case, you do not have single cell resolution. Brain regions are homogenized (so you lose visual resolution), but as a result, you are able to get expression data from all major gene products (~18,000ish) from a single sample, instead of 1 gene product per slice of brain (albeit a thin slice) - this is the case of the ISH technique they are currently using in the mouse brain.
@CMO: Thanks, that's all very helpful!
As for the connectome work by Lichtman and others - they are using much more sophisticated techniques. Not to discount the work of the Allen Inst, but those maps aren't really suited for doing this kind of work, although they may provide preliminary data to help design the more complex type of work.