Tunable multiphase dynamics of arginine and lysine liquid condensates

Abstract

Liquid phase separation into two or more coexisting phases has emerged as a new paradigm for understanding subcellular organization, prebiotic life, and the origins of disease. The design principles underlying biomolecular phase separation have the potential to drive the development of novel liquid-based organelles and therapeutics, however, an understanding of how individual molecules contribute to emergent material properties, and approaches to directly manipulate phase dynamics are lacking. Here, using microrheology, we demonstrate that droplets of poly-arginine coassembled with mono/polynucleotides have approximately 100 fold greater viscosity than comparable lysine droplets, both of which can be finer tuned by polymer length. We find that these amino acid-level differences can drive the formation of coexisting immiscible phases with tunable formation kinetics and can be further exploited to trigger the controlled release of droplet components. Together, this work provides a novel mechanism for leveraging sequence-level components in order to regulate droplet dynamics and multiphase coexistence.

Fig. 1. Viscosity of poly-L-lysine coacervates is controlled by polymer length.

a Brightfield image of polyK100/UTP condensates (10 mM Tris, pH 7.4). polyK concentration 6 mM per monomer and uridine triphosphate (UTP) 1.5 mM per monomer. b Widefield fluorescence image of polyK100/UTP condensate fusion (partitioned free Atto488 dye incorporated for enhanced visualization). c Confocal fluorescence image of polyK100/UTP droplet with 500 nm beads embedded (Red FluoSpheres, Invitrogen). Inset, representative 2D bead track. Scale bar 0.1 μm d Mean squared displacement (MSD) vs lag time for individual 500 nm beads in polyK100/UTP droplets. Inset, distribution of bead displacements at lag times = 0.5 s (red), 5 s (green), 10 s (blue). e MSD data for polyK10 (blue), polyK50 (red), and polyK100 (yellow) with UTP. Inset, viscosity as a function of polyK length. f Viscosity vs polyK length for polymers with UTP (blue), pU10 (green), and pU50 (yellow).

Fig. 2. Differences in assembly propensity of polyR and polyK droplets.

a DIC images showing (i) polyR10-UDP, polyR50-UDP, and polyR100-UDP (10 mM Tris, pH 7.4) with insets displaying polyK under same condition and (ii) polyR10-UTP, polyR50-UTP, and polyR100-UTP (10 mM Tris, pH 7.4). Concentration polyK/polyR 6 mM per monomer, UTP 1.5 mM, UDP 2 mM. Scale bar 20 μm. b Phase diagram for polyK50 (green) and polyR50 (purple) (6 mM per monomer) with varying NaCl and UTP concentrations. Green circles denote conditions under which polyK50-UTP droplet formation is observable. Purple circles denote conditions where polyR50-UTP droplet formation is observable.

Fig. 3. Differences in emergent properties of polyR and polyK droplets.

a Confocal fluorescence images of FRAP recovery for polyK10/pU10-A488 (upper) and polyR10/pU10-A488 (lower) illustrating increased fluidity of polyK vs polyR. b FRAP recovery within droplets of polyK/pU10-A488 (green) and polyR/pU10-A488 (purple). c MSD vs lag time for polyK and poly R of length 10 (blue, o), 50 (red, o) and 100 (yellow, o) with UTP) illustrating increased viscosity of polyR compared to polyK. Inset: Brownian motion of 200 nm bead in polyK50/UTP (green) and polyR50/ UTP (purple).

Fig. 4. Multiphase condensate behavior.

a–d Confocal fluorescence images of multiphase liquid condensates formed from the addition of UTP (1.5 mM (a), 3 mM, 4 mM or 15 mM (b–d)) to polyK:polyR 50:50 mixtures. Scale bar 20 μm e Confocal fluorescence images of fusion of dual-phase coacervates. polyK phase (green) and polyR phase unlabeled. f Aspect ratio change outer polyK droplet (green) and inner polyR droplet (purple). g Fusion timescale vs average droplet radius for polyR single-phase droplets (purple circles), polyR dual-phase droplets (purple crosses), polyK single-phase droplets (green circles), polyK dual-phase droplets (green crosses).

Fig. 5. Condensate inversion and polyK release.

a Confocal fluorescence images of polyK fluorescein isothiocyanate (FITC)-labeled (green) displacement by polyR50 labeled with dylight594 (purple). Merged images taken at moment of polyR addition (t = 0) and after 30, 60, and 90 s. Scale bar 20 μm. b Percentage of slide covered as a function of time for polyK (green) and polyR (purple). Inset images correspond to t = 0 and t = 90 s. Scale Bar 20 μm. c Close up of individual condensate green channel showing polyK FITC and purple channel only showing polyR50. Scale bar 5 μm. d Intensity of FITC-polyK over time inside a polyK droplet (filled square) and outside of a polyK droplet (open circle). Intensity values correspond to timeseries displayed in a. Inset displays polyK FITC before and after displacement by unlabeled polyR50 illustrating polyK displacement into surrounding media. Scale Bar 20 μm Intensity re-scaled in this image for clarity. [polyK] = [polyR] = 6 mM monomer. [UTP] = 1.5 mM.

Fig. 6. Control over inversion and multiphase coacervate creation.

a Confocal fluorescence images of droplet inversion via addition of Dylight-labeled polyR50 at increasing UTP concentrations (3, 4, 15 mM top to bottom). Timepoints 0, 100, 150, 200 s, and 1 h are shown. Scale bar = 20 μm. b Initial coacervates of polyK paired with (i) UTP, (ii) pU10, and (iii) pU50. Scale bar = 20 μm. c Intensity of FITC-labeled polyK in dilute phase for UTP concentrations 3 mM (red), 4 mM (gold), and 15 mM (blue). Intensity values correspond to timeseries displayed in a. d Intensity of FITC-labeled polyK in dilute phase for UTP (gold), pU10 (red), and pU50 (blue). Intensity values correspond to timeseries displayed in b.