Modeling stress granule dynamics: A quantitative biophysical framework linking condensate behavior to neurodegenerative pathogenesis

Abstract

Biomolecular condensates present a compelling frontier in biological engineering due to their dynamic material properties and ability to spatially and temporally organize biochemical reactions without membrane boundaries. These structures form through liquid-liquid phase separation, a thermodynamically driven process in which proteins and RNAs spontaneously form reversible, concentrated droplets. Stress granules (SGs) are a class of condensates that self-assemble in response to cellular stress and play a critical role in the regulation of mRNA triage during translational arrest. However, aberrant SGs have been implicated in the progression of neurodegenerative diseases such as Alzheimer’s, frontotemporal lobar degeneration, and amyotrophic lateral sclerosis, where persistent SG assemblies are associated with the pathological aggregation of TDP-43. Despite the growing recognition of this correlation, the biophysical mechanisms that drive SGs toward disease-relevant states remain poorly understood. To address this knowledge gap, this thesis presents a coarse-grained, reaction-diffusion model that integrates phase separation theory and reaction kinetics to investigate SG formation and aging during different stress regimes. The model tracks the interplay between SG components, including mRNA and RNA-binding proteins, and aggregation-prone proteins such as TDP-43. By using numerical computing to systematically vary key biophysical parameters, including interaction strength, cytoplasmic mRNA levels, and mRNP dissolution rates, we simulate how SG morphology, domain size, and composition evolve over time. Our model provides a quantitative and mechanistic framework for exploring how SGs shift from functional condensates to sites for fibril formation, offering insights into the physical principles linking phase separation to protein aggregation in neurodegenerative diseases.