Cells couldn’t have a hope of working if they weren’t tightly spatially organized. The nucleus vs. the cytosol (and the cell membrane itself) are the two most obvious partitions, and then you have specialized organelles like the mitochondria, et very much cetera, dividing things further. Life itself is organized around things being different on one side of a membrane as compared to the other.
Consider what happens when DNA (double-stranded) shows up in the cytosol, as opposed to its homeland inside the nuclear membrane. It’s just not supposed to be there, and its presence gets interpreted as a sign of infectious attack, triggering the innate immune system. That’s fine, most of the time, but there are complications with the immune system as a whole: people with some autoimmune diseases (the classic example is lupus) often display antibodies to dsDNA, probably inappropriate ones, and one theory is that there’s some sort of defect in apoptosis that allows leakage of (incompletely destroyed) endogenous dsDNA to the point that it raises an immune response. In general, you probably don’t want to set the immune system loose on your own genome – or on anything endogenous, for that matter.
So how do cells pick up on wandering bits of cytosolic DNA at all? In recent years, that’s become more clear. There’s a nucleotidyltransferase enzyme called cGAS (cyclic GMP-AMP synthase) that binds to cytosolic DNA and immediately starts synthesizing cGAMP when it does. That’s a second messenger that in turn activates another protein called STING (stimulator of interferon genes), which is involved in a host of cellular responses to trouble. This whole system is of great interest in drug discovery, naturally, with people in immunology, infectious disease, and cancer all piling into the area.
There’s a new paper that illuminates how all this works (commentary here) and guess what? It’s condensate time again. Previous work had already shown that cGAS/DNA complexes were organized into oligomers, but it now looks like these things come out of cytosolic solution entirely into small phase-separated droplets. These contain activated cGAS, and can combine with other such droplets as they come into proximity. ATP, GTP, and cGAMP rapidly diffuse in (and out) of the separate phase, and the degree of cGAS activity correlates with the amount of phase separation. The high concentrations in the new phase enhances the enzyme activity, apparently. The degree of phase separation, in turn, depends on the length of the dsDNA (a hot area of research all its own). Another factor in the ease of condensate formation is the concentration of zinc ions, and the authors suggest that this might be a regulatory mechanism for immune sensitivity. Most interestingly, RNA and single-stranded DNA can also cause a phase separation, but the cGAS in those droplets is not active.
This (like a lot of other recent condensate work) raises a whole host of new questions while it answer some others. Are there different sorts of cGAS droplets (other than dsDNA-active versus RNA/ssDNA inactive ones?) Is that zinc ion idea correct, and are there other cytosolic regulators of condensate formation? Do problems with these determine sensitivity to autoimmune responses? Even bigger: what other cytosolic enzymes work this way? Do such phase droplets, under some conditions, come down and stick to the surfaces inside the cell (the ER, nuclear membrane, mitrochondrial membanes, inner cell membrane), perhaps piling up concentrations of signaling molecules against receptors, or altering membrane properties in general?
More and more, I think it’s possible that we’re in the middle of rewriting the cell biology textbooks, so strap in. There will be a lot of crappy papers published in this area along with the good ones, to be sure, and claims will be made that don’t hold up. The hotness and trendiness will get on people’s nerves, and there will be fights for recognition and credit. But all of this is science. And there’s a lot of science going on around this idea right now.