Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical Sciences


Cancer and Developmental Biology

Research Advisor

J. Paul Taylor, PhD


Mondira Kundu, PhD; Tanja Mittag, PhD; Joseph T. Opferman, PhD; Tiffany N. Seagroves, PhD


Biomolecular condensation;Phase separation;SLiM;Stress granules


Biomolecular condensation has been recognized as an important strategy for spatial or- ganization within cells. Biomolecules condense through interactions. When the sum and duration of collective interactions reach the percolation threshold, a system-spanning inter- action network is achieved, resulting in phase separation to create a low-viscosity liquid, i.e., biomolecular condensate. Protein-protein interactions (PPIs) are essential for the formation of system-spanning interaction networks that drive biomolecular condensation. Short linear motifs (SLiMs) are the most abundant functional unit involved in protein-protein interactions (PPIs). The interactions between SLiMs and date hubs are crucial for various cellular processes. While SLiMs and date hubs are prevalent in biomolecular condensates and exhibit interaction attributes similar to those in the biomolecular condensate interaction network, the roles of the interactions between SLiMs and date hubs in contributing to biomolecular condensation remains unillustrated. Here, we take a prototypical biomolecu- lar condensate, stress granule (SG), as an example to illustrate how the interactions between SLiMs and date hubs contribute to the interaction networks that drive biomolecular con-densation.

We firstly confirmed that the NTF2L domain of G3BP1 protein, the central node of the SG interaction network, is a date hub that interacts with SLiMs. We found that the NTF2L domain has one binding surface, comprising four closely neighboring, degenerated binding areas. We then defined the SLiM patterns for NTF2L recognition and applied these SLiM patterns to predict novel NTF2L binding motifs in human proteome. We successfully verified that NUFIP2 is a novel NTF2L direct interactor. Leveraging the structural insights into the interactions between NTF2L and its binding SLiMs, we generated in silico NTF2L mutants designed to manipulate these interactions. We confirmed that two mutants, F33W and H31A, indeed altered the NTF2L interactome as predicted. Furthermore, we found that the interactions between NTF2L and SLiMs not only introduce NTF2L direct interactors into the SG interaction network, but these SLiM-containing proteins also serve as bridges, connecting many other SG components into the SG interaction network. Caprin1 bridges NTF2L and small ribosomal proteins, NUFIP2 bridges NTF2L and both ATXN2L and DDX6. The mutations in NTF2L impact recruitment of both NTF2L direct interactors and the indirect interactors dependent on the corresponding direct interactors. NTF2L F33W lost interactions with multiple direct interactors, including caprin1, USP10, and NUFIP2, and we showed that G3BP1 F33W impairs the assembly of SGs. Using SG as an example, we demonstrated that the interactions between SLiMs and date hubs play important roles in the interaction networks that drive biomolecular condensation. Manipulating the interactions between SLiMs and date hubs regulates biomolecular condensation.

Declaration of Authorship

Declaration of Authorship is included in the supplemental files.





Available for download on Friday, December 05, 2025