Protein aggregation is a real problem in the living cell. It’s become clear over the years that some proteins in particular are more prone to this behavior than others, with particular sequence motifs sometimes being signs of incipient trouble. But like much else in biology, there’s a trade-off: the same properties that make a protein more likely to self-associate can apparently make it useful in a variety of protein-protein interactions.
So cells have evolved ways of dealing with the problem. Chaperone proteins (such as the chaperonin TCP-1 ring complex, TRiC) try to start proteins out in the right conformations and keep things folded in useful (and non-dangerous) manners, but when aggregates inevitably form, the protein-destruction machinery kicks in. There are several of these degradation pathways – probably the most famous one is the decoration of proteins with ubiquitin chains, which mark them for consumption in proteasomes. Those are roughly tube-shaped structures in the cell (found both in the cytoplasm and the nucleus) that function as protein shredders. When a protein is delivered to one, there’s a gatekeeping process to make sure that yes, this one’s marked for demolition, then then ubiquitinated protein is fed into a series of high-capacity protease enzymes that break it into fragments for re-use.
There’s also autophagy, which feeds proteins into lysosomal compartments on a similar one-way trip. Several varieties of this process operate in cells, such as microautophagy, macroautophagy, mitophagy, xenophagy, ER-phagy, and chaperone-mediated autophagy. Some of these (like mitophagy and ER-phagy, don’t go through ubiquitination, but rather recognize specific sequences associated with particular organelles. Macroautophagy is another ubiquitination-driven process, going through structures called autophagosomes, and needing so-called “cargo receptor” proteins that recognize misfolded or aggregated proteins that have become ubiquitinated. p62 is a well-known member of that class, and there are quite a few others. All of them seem to bring in a key partner, autophagy-related protein 8s (ATG8s), that mediates the clearance of their targets.
Meanwhile, chaperone-mediated degradation applies to proteins with particular motifs (such as a KFERQ-like sequence), which is recognized by particular chaperone proteins that take them to the lysosomal membrane and start a process of translocation out of the cytosol. Interestingly, tumor cells often show high levels of chaperon-mediated degradation – it’s thought that they are removing proteins damaged by the aberrant cellular environment, or perhaps even clearing out tumor-suppressing proteins themselves.
Autophagy and proteasomal degradation are complementary and interconnected, and we’re still working out the ways that they fit together. One of the branches in the whole protein-degradation process for aggregates seems to be whether the aggregated proteins are still in a relatively liquid state (as in biomolecular condensate droplet liquid state), or have turned into solid chunks. It’s widely believed that “aging” of some of these liquid droplets takes them through a gel-like state and eventually to solid aggregates. p62, NBR1, and TAX1BP1 (among others) seem to promote autophagy of the more liquid varieties of ubiquitinated proteins, and this new paper names CCT2 as one that handles solidified aggregates – apparently the first such protein that has been discovered in higher organisms.
CCT2 also doesn’t seem to recognize ubiquitinated motifs, and apparently doesn’t cross paths with either macroautophagy or chaperone-mediated autophagy. The paper reports that CCT2 has both chaperone and “aggrephagy” functions, which are mediated by switching from a dimer to a monomeric form. The monomer exposes a motif that brings in ATG8 subtypes (binding to them by some unusual sequences which had not been observed before) and kicks off autophagy. It appears to associate with aggregation-prone proteins in general, probably acting as a chaperone to keep them in line, and switching to a destruct signal if things don’t hold up. Indeed, CCT2 seems to associate with that TRiC chaperone complex mentioned above when it’s a dimer, but dissociates as the monomer as it starts calling in the ATG8 proteins. Trying to refold troublesome proteins is the first line of defense, but autophagy is always there as an option should that fail or become overwhelmed.
There’s a type of retinopathy (leber congenital amaurosis, LCA) that’s been associated with point mutations in CCT2, and this paper proposes that these mutations disrupt the protein’s autophagy role more than they do its chaperone activity. Other lines of evidence suggest that protein aggregation is involved in LCA, and this might be the direct connection. Several neurodegenerative diseases are of course also well-known to be associated with protein aggregates, and CCT2 levels were shown several years ago to be reduced in brain tissue from Huntington’s and Alzheimer’s patients. It would be very much worth knowing if loss of CCT2 function is part of the etiology of such diseases – one way to check, as the authors note, would be to see if LCA retinopathy patients are more prone to neurodegeneration.
Just knowing that solid aggregates can indeed be cleared, and that there is a specific pathway for doing so, is a real advance. Are there ways to restore diminished CCT2 function, either at the protein express level or somewhere downstream? Therapy for these diseases needs all the good ideas that we can throw at them, and I feel sure that this will be put to the test as soon as someone figures out a good way to do it!
