Since 2012 the McKnight lab has studied an enigmatic class of protein domains specified by low complexity (LC) sequences. Certain LC sequences assemble into amyloid-like polymers, leading to formation of liquid-like droplets that sequentially mature into hydrogels. Although morphologically indistinguishable from pathogenic, prion-like amyloids, polymers formed from LC sequences disassemble upon dilution. Atomic resolution of the structure of FUS polymers has recently revealed the basis of LC domain polymer lability (11). In the six years since publication of our back-to-back 2012 papers in Cell (9,10), studies of RNA granules, membrane-less organelles and LC domains have exploded. That LC domains are involved in the formation of meso-scaled puncta not surrounded by investing membranes has been firmly established. Our observations favoring the involvement of labile cross-? interactions in the formation of these structures, however, stands in contrast to what has become the prevailing view in the field. Numerous groups have argued that LC sequences adopt no molecular structure upon phase transition. The McKnight lab has provided multiple lines of evidence showing that formation of labile cross-? interactions is at the heart of phase separation into liquid-like droplets, hydrogel formation, and LC domain function in living cells. Evidence supportive of these conclusions includes chemistry (14,16), pharmacology (12,16), correlative mutagenesis (10, 13, 14), human genetics (12, 15), and ? most recently - solid state NMR spectroscopy (11). The narrative of this application avoids this controversy and instead focuses on several objectives important to the advancement of our research. An assessment of differences between the McKnight perspective (LC domain function as driven by defined structural organization) and that of most others who have joined the field (LC domain function sans molecular structure) can be found in our review chapter recently published the Annual Review of Biochemistry (17). The two projects described in this application represent extensions of our work on the LC domain of FUS. We are also interested in other settings wherein cells employ labile cross-? interactions to abet normal or abnormal biology. We have found that the LC domains of intermediate filament proteins utilize labile, cross-? interactions to mediate filament maturation. In the context of assembled filaments, we have shown that coalesced LC domains represent binding sites for RNA granules (12). We also deduced the basis of C9orf72 pathogenesis via the toxic GRn and PRn poly-dipeptides produced by RAN translation of repeat expansion transcripts ? thus resolving how neurons die in the most prevalent form of ALS (12,15,16). Aside from continuing our collaborative SS-NMR studies with Dr. Robert Tycko, and NAI footprinting studies with Dr. Yonghao Yu to probe LC domain conformation in living cells, we have established a collaborative partnership with Dr. Ralf Langen of the University of Southern California to use electron paramagnetic resonance (EPR) methods for studies probing the earliest events wherein LC domains coalesce.
For over a century it has been known that eukaryotic cells contain a variety of basophilic puncta not invested by surrounding membranes. Cytosolic versions of these puncta include various forms of RNA granules, whereas nuclear versions have been described as Cajal bodies, nuclear speckles and PML bodies. The underlying science of this grant application offers a structural framework for probing how these puncta form, and how the dynamics of this enigmatic example of cellular organization may impact upon information flow from gene to mRNA to protein.
