A relatively recent study showed that transcription factors tend to hang out in specific areas of the nucleus, clustering like stars in galaxies. And it’s been rattling around in my head as potentially perfect for Mad Art Lab because they do, in fact, cluster like stars in galaxies — the statistical methods used to model both phenomena are exactly the same. It’s one of those rare times, with stars and life, when a technical explanation and a poetic one come out to the same thing.
What They Asked
Let’s start pretty basic: how does a cell know which genes to use? Some genes are “always on”, some pieces of DNA are always turned off, but most genes are sometimes on and sometimes off. One of the ways that they get turned on or off is by association with certain proteins, called transcription factors.
So how does a transcription factor find its gene, or genes, when they are scattered through millions and millions of bases of genome? For a while we’ve thought that basically, it’s a feat of random diffusion, with transcription factors floating through the nucleus, bumping up against various pieces of DNA, until they find the right one, where they sit for a while. And since these sites happen thousands of times in the genome, there are a lot of right answers.
But in terms of a search problem, phrased like that, it’s like finding a hundred hay-like needles within a needly haystack. It seems like a miracle transcription factors find their genes at all.
And they do; usually it takes a transcription factor about six minutes of searching to find a stable site. How does that work?
What They Found
The researchers used new methods to trace individual transcription factor molecules in live cells, and observed their behavior. What they found was this: throughout most of the nucleus, transcription factors diffused rapidly. But in certain regions, they bound — and when they were nearby, they slowed way down. It suggested that the binding sites — the genes that needed to be found — were clustering together.
So instead of a thousand hay-shaped needles spread evenly through the haystack, it’s a clump of five or ten or twenty. Which would be easier to find, and once you found it, you’re a longer way towards your goal of finding all 100.
It’s a pretty cool finding in its own right (in my biased opinion). But the thing that struck me most in terms this blog was that the test they used to model clustering in transcription factors was the same test generally used in astrophysics. They were, basically, asking to what extent the clusters of transcription factors looked like stars in a galaxy.
And the answer they got was “pretty much, yeah.” (at least as far as that can be a scientific answer).
The specific question is: given that I am a transcription factor (or a star), how likely is it that another transcription factor (or star) will be radius R away from me?
If there’s clustering, then smaller Rs will have bigger likelihoods. If there isn’t clustering, then it won’t matter the R. And for both transcription factors and stars, R matters a lot.
We Are Stardust
I think physics is full of metaphors because so much of it deals with things that are too big or too small to easily comprehend. But it means that many of the most beautiful science-metaphors that we encounter on a day to day basis are physics-related. And the best ones — “we are made from stars” — are true from a scientific sense (the atoms in our bodies were forged in long-dead stars) as well as a poetic/inspirational one.
To me, studies like this one just add yet another interpretation to the metaphor, pointing out that the rules and patterns that govern the organization of the universe are the same ones that govern the organization and growth of every living thing. It’s almost a spiritual thing, the sense that there is an order underlying everything. Except instead of being some ineffable deity’s will, it’s a rule we can learn, and use. And that’s pretty damn beautiful.
Featured image courtesy of Bildagentur Zoonar GmbH, shutterstock.