A new class of alanine-rich motifs that mediate liquid-liquid phase separation in RNA-binding proteins

The discovery that some cellular structures are liquid-like droplets and function as membrane-less organelles (MLO), was a major milestone in biology. The fact that they form spontaneously via liquid-liquid phase separation (LLPS), a well-known phenomenon from physical chemistry, has sparked intense interest into elucidating biophysical and molecular details that drive this process. This is important not only in terms of our deeper understanding of the sequence-structure-function relationship in proteins, but also because malfunction in LLPS has been identified as the cause of several human diseases (ALS, dementia, gene expansions etc). For this reason, there is a growing interest in understanding LLPS at the molecular level and finding ways to regulate this process using external factors, for example with small molecules. Biomolecular condensates are typically enriched in proteins with long disordered sequences containing low-complexity regions. These regions contain only several amino acid types, are repetitive and lack a stable secondary structure. Low-complexity regions can mediate LLPS in many proteins, often with other biomolecules, particularly RNA. A number of LLPS-promoting low-complexity sequence motifs have been identified and understanding how they contribute to LLPS is an area of intense research. These motifs can have different chemical 'characters': polar (enriched in G/S), positively charged (RG), can contain prion-like features (Q/N), or are more hydrophobic (V/P). They are frequently decorated with aromatic residues (Y/F). The current understanding is that these motifs mediate weak, short-ranged, multivalent interactions, through sticky residues (aromatics, charged) while polar residues maintain the liquid-like character of the condensate. I recent years sporadic evidence began to emerge which points to another, largely overlooked low-complexity motif which is enriched in alanine (AAAA). Alanine-rich sequences are known in the context of biomaterials such as spider silk and elastin, where they promote self-association and give unique material properties. Very recently such alanine-motifs were identified in two RNA-binding proteins (TDP-43 and Musasi-1) and were shown to mediate LLPS, most likely via a different mechanism based on the stable secondary structure elements. However, our preliminary data indicates that alanine-rich motifs are ubiquitous among RNA-binding proteins and likely represent a novel functional motif that promotes LLPS. Furthermore, alanine repeat expansions are a less-known group of rare genetic diseases and it has only recently been shown that these expansions can also alter the phase-separations in transcription factors. This leads to an open question whether alanine-rich sequences truly represent a novel class of functional motifs which promote LLPS. Given that alanine is a structure promoting amino acid with high propensity for helical structure, the mechanism of phase-separation seems to differ from the currently proposed models. The objective of this proposal is to investigate whether alanine-rich regions in RNA-binding proteins represent functional motifs which promote LLPS and to understand the biophysics behind it. We propose to use and interdisciplinary approach by combining bioinformatics, in vitro biochemical and biophysical methods and cell biology in order to: i) establish alanine-rich regions as bona fide LLPS-promoting motifs, ii) understand the biophysics of the phase-separation via alanine motifs and iii) develop a prediction tool to detect LLPS-promoting alanine-rich motifs in proteins. The results from this project will thus enable identification of new proteins undergoing LLPS, which will uncover new biology and open up new venues to modulate LLPS using small molecules. More broadly, elucidating what alanine rich sequences can do in terms of phase separation might prove crucial for understanding the molecular basis of polyalanine expansion diseases.