I’ve written here about phase-separated condensates inside cells, and the publications on these continue to show up all over the literature. I found this recent one in Science to be particularly interesting, on several levels.
One thing the condensate idea has given a framework to is the variety of small cellular structures that have been observed and named over the years, but without (in many cases) much idea of what they are. So you have things like nuclear speckles and paraspeckles and stress granules and interchromatin granules and PML bodies and Cajal bodies and P-bodies and germ granules. . .and on and on. It’s become apparent over the years that these are liquid phase-separated droplets, carefully thermodynamically balanced so that they can form and dissipate when needed. I am pleased to be able to link to a Science paper from 1899 where Edmund Beecher Wilson of Columbia noticed these liquid droplets and commented on their nature:
If the eggs of Ophiura be crushed by pressure on the cover-glass the protoplasm flows out, most of the alveolar spheres going in advance, while the granules and continuous substance lag behind. Meanwhile, the alveolar spheres often run together to form larger drops of all sizes, the origin of which is placed beyond question by their color. The same is true of the yellow microsomes, though this takes place less readily, and only under somewhat rough treatment. This demonstrates the liquid, or at least viscid, nature of both the spheres and the microsomes, and no less certainly that of the continuous substance in which both lie. . .
The most prominent of these structures is surely the nucleolus, a relatively large concentrated region inside the nucleus whose liquid nature had also been noted over the years. At right is a shot from Wikipedia, which also shows the membraneless nature of the structure (as opposed to the nucleus itself, whose membrane is clearly visible). Why exactly there is a nucleolus is another question. And it’s one of those that tends to be rather hard to answer – as is usual in science, you can find a vast number of papers on its behavior under different conditions, on its composition (at increasingly fine levels of discrimination), etc. But questions such as “What does it do?” or “Why is it there?”, while very easy to ask, often have to wait for a large accumulation of such data before being answered. Not that a large accumulation of data guarantees anything of the kind.
You can see from that photo that the nucleolus itself has internal structure, and that’s been the subject of a lot of work as well. It’s known (for example) that the fibrillar center is where ribosomal RNA gets transcribed, and ribosomal subunit assembly is another function. But there are a vast number of protein-RNA-DNA interactions going on in there (involving a very long list of different proteins), so many aspects of nucleolar function are still a mystery.
We may have another part of the answer, though. The new paper linked above (from teams at Max Planck/Martinsried and the Ludwig Maximilian U. in Munich) shows that at least one function (out in the granular compartment) is as an emergency refuge for nuclear proteins during stress. They associated with other proteins such as NPM1 to stabilize their conformations and keep them from aggregating, and then when conditions improve, chaperones such as HSP70 help extricate and refold them. The liquid nature of the GC region seems to be a key feature in keeping the proteins from further damage, and it only has so much capacity. Under prolonged stress, it apparently gets overloaded, and eventually goes from a liquid state to more of a stiff gel (which behavior has been noted in other condensate structures, both in vitro and in vivo). In this case, that transition seems to be irreversible – HSP70 et al. can’t bring things back, and my guess would be that if the cell survives such treatment at all that the solidified material eventually gets dragged off for some sort of degradation and recycling.
The group added a nuclear localization signal peptide region to firefly luciferase and watched it distribute through the nucleus. Under heat stress the protein distributed to the GC region of the nucleolus, and it apparently had to be unfolded for this to happen (if they added a tight-binding luciferase inhibitor, this impaired the redistribution under heat). About 60% of the protein didn’t make it back once the heat shock period had ended, but the 40% that did went back out into the nucleoplasm and was active again. Interfering with HSP70 activity, though, blocked this process.
So the nucleolus is a sort of lager for nuclear proteins in times of stress – considering the amount of stress involved, maybe that should be the Afrikaans laager instead (a very old tactic, as that link shows). In this quality control system, they’re preserved in the liquid-phase region in a form that allows them to be refolded later on (or, presumably, to be easily ubiquitinated for the nuclear proteasomes to demolish them if refolding doesn’t work out). The alternative is for such proteins to aggregate into amyloid-type insoluble forms that can’t be easily refolded or degraded at all, and we’re seeing the machinery that’s grown up around them to keep this from happening. The paper identifies at least two hundred proteins that relocate in this fashion, including some that are considered to be mostly cytosolic, which is interesting. And it’s quite possible that a breakdown in this process is involved in some protein-misfolding diseases as well. . .