For upwards of 30 years it has been known that regulatory proteins involved in the control of gene expression and mRNA biogenesis contain low complexity (LC) polypeptide sequences that are intrinsically disordered. In the case of gene specific transcription factors, these LC sequences, often typified by long, homopolymeric segments of polyglutamine, have been implicated in the illusive process of transcriptional activation. It is well-understood how the DNA binding domains of transcription factors, including zinc fingers, leucine zippers, bHLH domains, and homeobox domains, facilitate direct and specific interaction between transcription factors and their target genes. By contrast, the field remains largely ignorant of the mechanisms by which the LC sequences typifying "activation domains" facilitate gene expression. In the case of proteins that regulate RNA biogenesis, similarly perplexing LC sequences are often linked to well-folded domains that mediate direct interaction with RNA substrates (including RRM, KH and pumlio domains). By serendipity, members of the McKnight laboratory discovered that a biotinylated isoxazole (b-isox) chemical is capable of reversibly co-precipitating hundreds of RNA binding proteins from cytosolic extracts. This unexpected discovery formed the starting point for experiments showing that LC sequences can reversibly polymerize into cross-beta filaments that phase transition to a hydrogel-like state. In turn, microscopic versions of these hydrogel droplets were evolved into a simplified, quantitative assay for gel retention of homotypic and heterotypic test proteins. By discovering that repeats of the sequence [G/S]Y[G/S] serve as a nucleating substrate for fiber polymerization, it was possible to test the correlative effects of mutational impediments to fiber polymerization in vitro with the ability of RNA binding proteins to move in and out of RNA granules in living cells. As a result of this work, it has been hypothesized that reversible polymerization of LC domains may represent the organizational basis for the formation of RNA granules including P-granules, stress granules, P- bodies and neuronal granules (Kato et al., 2012;Han et al., 2012). As an extension of this work nuclear extracts have been exposed to the b-isox chemical. These experiments led to the reversible precipitation of many nuclear proteins containing LC sequences, including TATA binding protein, TAF15, the largest subunit of RNA polymerase II, and components of the mediator complex. A major objective of the experiments proposed in this grant application will be the test of whether similar processes are employed to control the organization of nuclear puncta, including Cajal bodies, nuclear speckles and "transcription factories".
For over 100 years it has been known that eukaryotic cells contain dynamic 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 transcription factories. The underlying science of this grant application offers a structural framework for understanding how these puncta form, and how the dynamics of this enigmatic example of cellular organization may impact on information flow from gene to mRNA to protein.