RNA-driven phase transitions in biomolecular condensates

Abstract

RNAs and RNA-binding proteins can undergo spontaneous or active condensation into phase-separated liquid-like droplets. These condensates are cellular hubs for various physiological processes, and their dysregulation leads to diseases. Although RNAs are core components of many cellular condensates, the underlying molecular determinants for the formation, regulation, and function of ribonucleoprotein condensates have largely been studied from a protein-centric perspective. Here, we highlight recent developments in ribonucleoprotein condensate biology with a particular emphasis on RNA-driven phase transitions. We also present emerging future directions that might shed light on the role of RNA condensates in spatiotemporal regulation of cellular processes and inspire bioengineering of RNA-based therapeutics.

Keywords: RNA chaperones; RNP granules; gene regulation; helicases; liquid-liquid phase separation; percolation; stress granules.

Figure 1.. Select inventory of RNP granules.

Various RNA-containing granules, which can be either nuclear or cytoplasmic, cell-type, and/or organism-specific, are depicted. Each RNP granule together with key RNA components are listed, red- or blue-colored fonts are used to differentiate RNAs that play a role in disease versus cellular architecture, respectively. Although not covered in the text, we include here the paracrine granules, RNase L-dependent bodies, Q-bodies, mei2 dots, histone locus bodies, TNBL aggregates, omega speckle, HSATIII stress bodies, BRD3 bodies, COOLAIR, SHORTROOT, SPA2/BNI1, and nYACs. The mesoscale structure of RNP condensates can vary greatly (as depicted), depending on the contextual role of specific and non-specific RNAs in these condensates and the interactions with diverse protein components.

Figure 2.. RNA modulation of RNP condensate formation, size, and stability.

(A) A representation of the titration of RNA against an RBP. The addition of RNA enhances PS of the mixture by increasing the concentration of charge-neutral complexes above C_sat. Further increase in [RNA] shifts the system to a super-stoichiometric regime, where the concentration of the charge-neutral complexes falls below the C_sat and thereby inhibits PS. (B) RNA modulates the core-shell architectures of RNP condensates at non-stoichiometric conditions. At low [RNA], the shell phase is enriched in positively charged RBPs, while at high [RNA], the architecture and the surface charge are inverted^,. (C) Schematic of RNA acting as a scaffold for RBP PS. (D) RBP microphases can be stabilized by RNA acting as a surfactant.

Figure 3.. Comparison of PS coupled to percolation for different RNA systems.

(A) the two-phase coexistence line of the RNAs corresponds to a heat-induced LCST-type PS. The intersection between the phase-separated regime (2-Φ) and the percolation (prc) line defining a networking transition is context-dependent. The degree of coupling between PS and percolation determines the classification of the systems either as non-percolating (fully reversible), weakly percolating (partially reversible), or strongly percolating (irreversible). (B) Visual depiction of RNA phase transitions as classified in (A). An ensemble of single RNA chains is shown with varying degree of intra-molecular structures. The reversibility of RNA condensates is determined by the degree of intra-condensate network formation. (C) Weakly or non-percolating RNA do not form extensive percolated networks allowing condensates to behave as terminally viscous fluids. A fully percolated RNA condensate is dynamically arrested and behaves as a terminally elastic solid. These differences in solid vs. liquid material properties can be probed by measuring differences in the relationship between the applied stress and measured strain at the nano-scale.

Figure 4.. Passive and active modes of regulating RNA percolation.

(A) An energy landscape of single-chain folding and multivalent RNA-RNA interactions. The light red-colored gradient indicates increasing [RNA]. The intra-chain and inter-chain interactions are separated by the energetic barrier which is proportional to the depth of the folding energy of the monomers. The barrier between single-chain and clusters emphasizes that folding processes of single RNA chains are on a different scale than the diffusion-limited process of nucleation and the height of the energy barrier is arbitrary. (B) RBPs passively buffer intra-condensate percolation-driven by RNA-RNA interactions while RNA helicases use ATP to actively remodel intermolecular RNA-RNA interactions. (C) RBPs can employ spontaneous and ATP-driven mechanisms to modulate RNA percolation. (Left) Heterotypic interactions mediated by multivalent RBPs (purple) buffer homotypic RNA interactions leading to an increased energetic barrier to the percolated state. (Right) RNA helicases can actively fluidize percolated RNA condensates allowing dynamic network reconfiguration and leading to greater mobility of RNA inside percolated networks.

Figure 5.

(A) A schematic illustrating the concentration-dependent, self-limiting feedback mechanism of RNA in regulating transcription through reentrant liquid condensation. (B) The formation of RNP granules driven by spermiogenic mRNAs and increased concentrations of FXR1 leads to increased translation during spermiogenesis. (C) When there is an increased demand for glucose catabolism, glycolytic mRNAs are localized and translated within core fermentation granules which act as translation factories and increase production of glycolytic enzymes. (D) During early embryogenesis, maternal transcriptome is degraded and replaced by zygotic transcripts. The decay of maternal mRNAs is regulated by dynamic condensation and decondensation of FMR1 granules.