Non-associative phase separation in an evaporating droplet as a model for prebiotic compartmentalization

Abstract

The synthetic pathways of life's building blocks are envisaged to be through a series of complex prebiotic reactions and processes. However, the strategy to compartmentalize and concentrate biopolymers under prebiotic conditions remains elusive. Liquid-liquid phase separation is a mechanism by which membraneless organelles form inside cells, and has been hypothesized as a potential mechanism for prebiotic compartmentalization. Associative phase separation of oppositely charged species has been shown to partition RNA, but the strongly negative charge exhibited by RNA suggests that RNA-polycation interactions could inhibit RNA folding and its functioning inside the coacervates. Here, we present a prebiotically plausible pathway for non-associative phase separation within an evaporating all-aqueous sessile droplet. We quantitatively investigate the kinetic pathway of phase separation triggered by the non-uniform evaporation rate, together with the Marangoni flow-driven hydrodynamics inside the sessile droplet. With the ability to undergo liquid-liquid phase separation, the drying droplets provide a robust mechanism for formation of prebiotic membraneless compartments, as demonstrated by localization and storage of nucleic acids, in vitro transcription, as well as a three-fold enhancement of ribozyme activity. The compartmentalization mechanism illustrated in this model system is feasible on wet organophilic silica-rich surfaces during early molecular evolution.

Fig. 1. Evaporation-triggered segregative LLPS inside the all-aqueous sessile droplet.

a Schematic drawing of early genetic molecule compartmentalization inside the evaporating aqueous droplets. According to timeline of the early history of life, the Pre-RNA world and RNA world may have occurred at 3.8–4.0 Gya, with the first RNA polymers formed in warm little ponds. Silica-rich surface could serve as the reactor that supports various prebiotic reactions such as the polymerization of the polynucleotides;^, b, d Phase diagrams of PEG and dextran mixtures. The red solid line is the binodal curve that distinguishes the single-phase region and the two-phase coexistence region. The arrowed line represents the tie line of ATPS, along which a mixture undergoes liquid-liquid phase separation (LLPS) and forms a PEG-rich phase and a dextran-rich phase. The star represents the composition of the sessile droplet, with 5 wt% PEG–10 wt% dextran in (b) and 9 wt% PEG–4 wt% dextran in (d); c Phase-separated pattern evolution inside an evaporating droplet of regime 1, shown in both bright-field image sequence and fluorescence image sequence, respectively; The dextran-rich phase is labeled by fluorescein isothiocyanate-dextran (FITC-dextran, green) and PEG-rich phase is labeled by Rhodamine B (red). e Phase-separated pattern evolution inside an evaporating droplet of regime 2, shown in bright-field image sequence and fluorescence image sequence, respectively; The fluorescence labeling is the same with that in (c). For c and e, images are obtained over analysis of seven independent trials with relative humidity ranging from 55 to 65%. The scale bar is 500 μm.

Fig. 2. Phase separation dynamics inside the evaporating droplet.

a, b Schematic drawings for polymer self-organized patterns in regime 1 (a) and regime 2 (b), with the definition of LLPS front that distinct single-phase region (light green color) and phase separated region (light purple for PEG-rich phase and dark purple for dextran-rich phase) inside the sessile droplet, which are demonstrated by the red solid circles with a radius of $R_2$; c Phase diagram of evaporation-triggered polymer self-organization regimes inside the sessile droplet; The phase diagram is the result of observations and analysis of at least three independent trials for each data point, with relative humidity ranging from 55 to 65%. d The radius of LLPS front $R_2$ as a function of evaporation time. Error bars represent SEM (standard error of the mean) from three independent experiments; Insert: Schematic drawing of nonuniform evaporation rate induced Marangoni effects. A surface tension difference $\mathrm{\Delta }\gamma$ is generated due to the nonuniform evaporation flux along the droplet surface. e, f Kinetic pathway of LLPS in regime 1 (e) and regime 2 (f). A finite small area ($△r=0.05R_0$) near droplet edge is chosen as the calculation domain, with a time step (denoted by red arrows) of $△t=5s$; $I_0$ and $I_0^{\prime }$ are initial compositions of the single-phase droplet. $F_0$ and $F_0^{\prime }$ are compositions within the calculation domain after the first time step, with their exact values given in Supplementary Table 1. Once surpassing the binodal curve, the mixture spontaneously phase separates into a PEG-rich phase ($F_1$ and $F_1^{\prime }$) and a dextran-rich phase ($F_2$ and $F_2^{\prime }$), respectively. Similarly, $S_0$ and $S_0^{\prime }$ correspond to the compositions within the calculation domain after the second time step (Supplementary Table 1), with the formation of a new PEG-rich phase ($S_1$ and $S_1^{\prime }$) and a new dextran-rich phase ($S_2$ and $S_2^{\prime }$), respectively. Source data are provided as a Source Data file.

Fig. 3. Compartmentalization and localization of DNA inside the evaporating droplet.

DNA localizes into dextran-rich compartments inside the evaporating sessile droplet. DNA is labeled with Cy-5 (red) and dextran-rich compartments are labeled with FITC-dextran (green). The ATPS solution is made up of 9 wt% PEG and 4 wt% dextran, with the addition of Cy-5 labeled DNA (10 µM) and FITC-dextran ($M_w=4000,$ 0.5 wt%). Images are obtained over analysis of seven independent trials with relative humidity ranging from 55 to 65%. The scale bar is 500 μm.

Fig. 4. In vitro transcription of functional RNA aptamers inside the evaporating sessile droplets.

a The flow of genetic information in phase-separated compartments via in vitro transcription (IVT). The DNA templates are coded for Broccoli aptamer that can bind 5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI) to form a fluorescent complex of Broccoli–DFHBI. b Optical image sequence shows fluorescence RNA aptamer transcription from Broccoli DNA template inside the phase-separated compartments. The dextran-rich compartments have increasing green fluorescence due to the formation of Broccoli-DFHBI complexes. Images are obtained over analysis of three independent trials with relative humidity ranging from 55 to 65%. The scale bar is 500 μm.

Fig. 5. Enhanced ribozyme cleavage by compartmentalization.

a Schematic drawing of the RNA cleavage reaction. The fluorophore labeled hammerhead substrate (HHS) is cut into two smaller pieces by the hammerhead ribozyme (HHE) and emits green fluorescence. The substrate contains a donor that is quenched by FRET when HHS is not cleavage. b Image sequence shows RNA cleavage process inside the phase-separated compartments. c Control group of the RNA cleavage reaction inside an evaporating droplet while without HHE added. There is only fluorophore labeled substrate (HHS) introduced into the droplets. d Fluorescence intensity evolution of experimental group (both HHS and HHE are introduced into the evaporating ATPS droplet; Solid line), no ribozyme control group (only HHE is introduced into the evaporating ATPS droplet; Dashed line) as well as no ATPS control group (both HHS and HHE are introduced into the evaporating water droplet; Dash-dotted line). The intensity value is obtained by gel analysis of a specified rectangular area (as shown in (b) and (c)) in the fluorescence images. Images are obtained over analysis of three independent trials with relative humidity ranging from 55 to 65%. Insert: Dimensionless productivities $m_p$ of ribozyme cleavage in the domain of dextran-rich compartment (red solid line) and that of water droplet (gray solid line) as a function of dimensionless time $t^{\prime }$. Source data are provided as a Source Data file. The scale bar is 1 mm.