Understanding the sequence and structural determinants of phase behavior of ALS-causing proteins

Summary Amyotrophic lateral sclerosis (ALS) is a life-threatening, neurodegenerative disease that causes the degeneration of motor neurons in the brain and spinal cord. There are currently neither a cure nor effective treatments to slow progression. However, recent new genetic, biochemical and biophysical evidence implicates stress granules as crucibles for disease development. Stress granules are membraneless organelles, also called biomolecular condensates, which form via liquid-liquid phase separation (LLPS) of RNA-binding proteins and RNA. Mutations in RNA-binding proteins convert liquid-like stress granules into solid inclusions. Prolonged stress granule assembly can result in similar effects. These observations point to new opportunities for therapeutic interventions if key open questions regarding the nature of liquid vs. solid assemblies can be answered. We will thus test the overarching hypothesis, which is based on above observations, that mutations in RNA-binding proteins change the driving forces for phase separation, the dynamical arrest of the liquid condensates and the ability of the condensates to promote the formation of protein fibrils. Our proposed studies will thus focus on the physics of phase separation of RNA-binding proteins, specifically on their intrinsically disordered low-complexity domains (LCDs) that are sufficient for mediating phase separation and are the typical locations of disease mutations. We will use the LCD of hnRNPA1 as an archetypal member of the class of ALS-associated RNA- binding proteins and will extend our studies also to the LCD of FUS. Mittag and Pappu have recently developed a stickers-and-spacers model that is based on the identification of transient, cohesive interactions amongst aromatic amino acid residues as providing the main driving force for phase separation. The aromatic residues are the stickers in this model, the spacers are the residues that connect the stickers. The model enables the quantitative prediction of full coexistence curves as a function of temperature and, importantly, resulted in a conceptual advancement of our understanding of how phase separation is encoded in LCDs. The complimentary expertise of Mittag and Pappu will now bring to bear a combination of biophysical experiments, computation and theory on the following three specific aims: (1) To extend the stickers-and-spacers model by quantifying the interplay among different types of stickers and spacers. (2) To test the hypothesis that disease causing mutations within LCDs of ALS-causing RNA-binding proteins cause dynamically arrested phase transitions. (3) To uncover the interplay among sidechain and backbone interactions and their contributions to spatial organization of LCDs within dense phases. Our results will enable quantitative predictions of the effects of ALS-associated mutants on phase behavior. We will obtain a clear understanding of how sequence-specific phase diagrams contribute to the dynamics of phase separation and aging phenomena. We will identify the types of interactions underlying liquid-like and solid-like dense phases. These results will have a direct bearing on therapeutic interventions against the functional disruptions that are likely to be caused by dynamically arrested phase separation.