Project Summary Abstract Nuclear speckles are an archetypal biomolecular condensates that are essential to either activating or repressing target gene expression. Unsuccessful control of gene expression is associated with development and progression of human cancers. Understanding the mechanism underlying nuclear speckle formation and functions can uncover possible modes of action for therapeutics. The main scaffold of nuclear speckles is formed by RNA and RNA binding proteins (RBPs), which have modular architectures of multiple domains, termed RNA recognition motifs (RRMs) and intrinsically disordered regions (IDRs). I propose to develop a comprehensive approach to study the specific arrangement of RRMs and IDRs in RBP assemblies and how each of RRM or IDR affects nuclear speckle formation and functions. Typically, RRMs can be readily identified by protein sequence that is evolutionarily maintained and their functions can be predicted. However, function cannot be easily predicted for IDRs from their sequences due to a large variability throughout evolution. Using modern computational analyses, I have identified the distinct molecular architecture and sequence features of the IDRs in nuclear speckle proteins that are obscured by variability in sequence. These indicate that the proteins should undergo a specific arrangement into assemblies with limited size. I will test these predictions by experiments using isolated proteins and RNA, and in live cells. I will use model nuclear speckle proteins, the SRSFs. In Aim 1, I will experimentally test how variations in the molecular architecture contribute to nuclear speckle formation by studying SRSF family proteins that naturally vary in the number of RRMs and IDR sequence features. Aims 2 and 3 will be studied during the independent phase of this grant (R00). In Aim 2, I will investigate how the individual assemblies of SRSF proteins are held together by a core of speckles. In Aim 3, I will investigate how different nuclear speckle proteins are released and move outside of nucleus under cellular stress. This award will support training in computational simulations and advanced experiments. Together these aims will support the long-term goal of uncovering the molecular mechanism underlying assembly and functions of nuclear speckles from the distinct features in the protein sequence.