Molecular Determinants for the Layering and Coarsening of Biological Condensates

Abstract

Many membraneless organelles, or biological condensates, form through phase separation, and play key roles in signal sensing and transcriptional regulation. While the functional importance of these condensates has inspired many studies to characterize their stability and spatial organization, the underlying principles that dictate these emergent properties are still being uncovered. In this review, we examine recent work on biological condensates, especially multicomponent systems. We focus on connecting molecular factors such as binding energy, valency, and stoichiometry with the interfacial tension, explaining the nontrivial interior organization in many condensates. We further discuss mechanisms that arrest condensate coalescence by lowering the surface tension or introducing kinetic barriers to stabilize the multidroplet state.

Figure 1:

Examples of biological condensates with a layered organization. A, left) Schematic diagram of subcompartments within the nucleolus. A, right) Nucleoli within an untreated X. laevis nucleus. NPM1 (red), FIB1 (green) and POLR1E (blue) are tagged. Scale bar, 20 µm. Image modified from with permission from Elsevier. B) Organization of two components within nuclear speckles, the protein SON and the snRNA U2B”. Scale bar, 1 µm. Adapted with permission from the Journal of Cell Science. C) Photomicrographs examining the in vivo assembly of MEG-3:meGFP and PGL-3:mCherry, two main components of P granules. Scale bar, 500 nm. Image minimally modified from. Reprinted with permission from AAAS. D) Stochastic optical reconstruction microscopy (STORM) image of a stress granule (gray), highlighting poly(A+) RNA cores (yellow). Scale bar, 500 nm. Image modified from with permission from Elsevier.

Figure 2:

Relation between interfacial tension (τ), effective interaction parameter (χ), and condensate organization. A) Mixing two immiscible droplets can lead to four possible organizations: (1) a layered droplet with dropletA on the inside, (2) partially wetted droplets that share an interface (3) complete nonwetting to form two separate droplets, and (4) a layered droplet with dropletB on the inside. The most stable organization is determined by the interfacial tension between dropletA and dropletB (τ_AB), between dropletA and the solvent phase (τ_As), and between dropletB and the solvent phase (τ_Bs). The droplet size (α_A and α_B) can also contribute to the condensate organization, as demonstrated by Lu and Spruijt. B) For immiscible homopolymers, the Flory-Huggins interaction parameter (χ_A, χ_B) can provide a way to approximate differences in surface tension. C) Flory-Huggins theory can be generalized to heteropolymers by assuming an effective interaction parameter that averages over differences in the sequence (${\chi }_A^{\text{eff}}$, ${\chi }_B^{\text{eff}}$). D) Heteropolymers can be divided into segments with different physical properties, resulting in effective parameters for different portions of the single chain (${\chi }_A^{\text{1}}$, ${\chi }_B^1$).

Figure 3:

Molecular factors that drive the formation of layered condensates. (A) Hypothetical interaction patterns of polyR, polyK, and UTP that explain the observed condensate organization. (B) Confocal fluorescence images of 50:50 polyK(green):polyR(purple) mixtures at different ratios of UTP. Scale bar, 20 µm. (C) Confocal fluorescence images of fusion of layered coacervates. polyK is labeled in green, and polyR is unlabeled. Images modified from ref. CC BY 4.0. (D) Approximate interaction patterns of HP1α, H1, and DNA, which also results in a layered condensate. (E) The slab density profiles support a layered organization for mixtures of HP1α (blue), H1 (green), and DNA (red). HP1α coalesces toward the center of the droplet, with H1 to the outside. Image modified from with permission from Elsevier. (F) Sequence diagram of the native LAF1-RGG sequence (RGG) and the shuffled RGG sequence (RGG_Cshuf). Anonic (red) and cationic (blue) amino acids are highlighted. SCD is a measure of charge patterning, where larger, negative values indicate segragated regions of the same charge. (G) Potential of mean force (PMF) for protein-protein and protein-RNA interactions with RGG and RGG_Cshuf. (H) Simulation snapshots indicating the preference of A₁₅ to localize to the condensate exterior with RGG_Cshuf but not RGG. (I) Density profiles of A₁₅ in RGG and RGG_Cshuf condensates. Reprinted from with permission from Oxford University Press.

Figure 4:

Molecular factors that limit condensate growth. (A) Computational evidence that lower valency surfactants can limit droplet growth. Surface tension (ρ) depends on the ratio of surfactants (red) to scaffold (blue). Vertical dashed lines indicate the maximum surfactant concentration that allows for phase separation for a given number-droplet regime. Note that the maximum droplet size varies continuously with surfactant concentration even within the same number-droplet regime. Snapshots of simulations in each droplet regime are included. Images modified from ref. CC BY 4.0. (B) RNA binding modifies ArtiG size. Confocal images of ArtiG in HeLa cells, 24 hours after transfection of mCherry-FFm and PUM.HD-FFm constructs at ratios of 1:1 (i), 5:1 (ii), 10:1 (iii), and 1:0 (iv). ArtiG^mCh indicates ArtiG comprised of mCherry-FFm, while ArtiG^mCh/PUM indicates ArtiG comprised of both mCherry-FFm and PUM.HD-FFm. Images modified from ref. CC BY 4.0. (C) Computational modeling shows the presence of an entropic barrier that stabilizes the two-droplet state for nucleoli. Free energy profile as a function of the radius of gyration, which effectively measures the distance between the two droplets. The free energy is broken into entropic (red) and energetic (black) components before (D) and after (E) the barrier for droplet fusion. Images modified from ref. CC BY 4.0.