Active coacervate droplets as a model for membraneless organelles and protocells

The study addresses how membraneless organelles, which are liquid–liquid phase-separated condensates of proteins and RNA, regulate their dynamic composition via ATP-dependent reaction cycles. While numerous coacervate models exist, most transition between equilibrium states in response to environmental changes, leaving unclear how chemically fueled, out-of-equilibrium processes control assembly, composition, and dynamics. The authors propose a fuel-driven coacervate droplet system as a model to mimic biological regulation of condensates, aiming to uncover how reaction kinetics govern droplet formation, growth, dissolution, composition, and functional RNA partitioning—phenomena relevant both to intracellular organelles and protocell models.

Prior work established that membraneless organelles form via complex coacervation and can be regulated by ATP-dependent cycles. Synthetic coacervate systems respond to pH, temperature, salt/polymer concentration, light, or enzymatic reactions, but these generally move between equilibrium states. The field of dissipative self-assembly shows that chemical reaction cycles with activation and deactivation steps can transiently form structures (fibers, vesicles) maintained by fuel consumption. Theoretical proposals suggest chemically driven droplets can grow and divide as protocells. Building on this, the present work aims to realize a chemically fueled, RNA-containing coacervate model to interrogate dynamic regulation, composition control, and potential division-like behaviors.

Design: The droplet system uses polyuridylic acid (poly-U, ~2,200 bases) as RNA and a designed peptide Ac-FRGRGRGD-OH. A carbodiimide fuel (EDC) activates the C-terminal aspartate of the peptide to a transient anhydride, increasing net cationic charge (precursor overall +1; product +3), thereby enhancing RNA affinity and inducing complex coacervation. The product spontaneously hydrolyzes back to the precursor, completing the fuel-driven activation/deactivation cycle. Experimental conditions: Typical mixtures contained 23 mM peptide precursor, 4.1 mM RNA (uridine units), 200 mM MES buffer, pH 5.3. Fuel concentrations varied (e.g., 25 mM for dynamic droplets; 60 mM for metastable droplets). Assays and instrumentation: Turbidity at 600 nm monitored droplet presence and lifetime via plate reader. HPLC quantified fuel (EDC), waste (EDU), and indirectly quantified anhydride product via benzylamine trapping to a mono-amide (due to anhydride instability). A kinetic MATLAB model (ODEs for EDC, precursor, O-acyl urea, anhydride, EDU) fitted rate constants to reproduce measured kinetics. Isothermal titration calorimetry (ITC) measured precursor–RNA binding (Ka). Confocal fluorescence microscopy (Cy3-/Cy5-labeled A15 and functional RNAs) imaged droplet formation, fusion, vacuole formation, fragmentation, and dissolution over time; image analysis quantified droplet number, volume, and fluorescence intensity. Phase partition analyses used filtration (0.45 µm regenerated cellulose) and centrifugation to separate droplet and supernatant phases; fluorescent RNA quantitation in supernatant assessed RNA capture. Broccoli aptamer activity with DFHBIT ligand was tested by fluorescence spectroscopy and confocal microscopy under low to moderate Mg2+/K+ to confirm folding and function inside droplets. Control experiments verified that fuel, RNA, and peptide were all required for droplet formation, and that droplets could be re-formed by refueling. Parameter sweeps across fuel and RNA concentrations built a phase diagram delineating no droplet, dynamic, and metastable regimes.

- Fuel-driven emergence and decay: Adding EDC rapidly formed droplets (turbidity rise within minutes) that decayed as fuel/product were consumed; without fuel, no droplets formed. Dynamic droplets (e.g., 25 mM EDC) exhibited turbidity peaking at ~3 min and returning to baseline by ~18 min, correlating with the transient anhydride product profile. - Metastable regime: At higher fuel (e.g., 60 mM EDC), droplets persisted (turbidity for ~76 min) despite fuel and product depletion by ~30 min, indicating kinetic arrest during disassembly. - Critical coacervation concentration: A minimum of ~0.9 mM anhydride product (achieved with >7.5 mM fuel) was required to induce turbidity (droplet formation). - Phase diagram: 7.5–40 mM fuel produced dynamic droplets; >40 mM fuel produced metastable droplets. Lower RNA concentrations shifted the threshold for metastability (e.g., metastable at ≥25 mM fuel with 1.4 mM RNA). - Partitioning of components: Fuel resided predominantly outside droplets throughout. In contrast, a significant fraction of anhydride product partitioned into droplets early in the cycle (e.g., after 2 min in dynamic droplets, ~1.7 mM product remained in filtrate, ~55% of total, indicating substantial intra-droplet product). Calculated intra-droplet product concentrations after 2 min were ~0.9 M (dynamic) and ~1.8 M (metastable), implying denser packing and a higher cation:anion ratio (2.0 metastable vs 1.1 dynamic). Metastable droplets captured essentially all RNA from solution; dynamic droplets left ~10% RNA in supernatant (measured 2 min post fueling with 25 mM fuel). - Microscopy dynamics: Dynamic droplets exhibited rapid nucleation and fusion early, followed by vacuole formation (>12 min) and terminal fragmentation/dissolution (~14–22 min). Fragmentation produced smaller non-spherical segments that could transiently persist. Metastable droplets retained RNA fluorescence intensity over time, consistent with RNA retention even post product hydrolysis. - Functional RNA up-concentration: Functional RNAs (SunY ribozyme 187 nt, Hammerhead ribozyme 44 nt, Broccoli aptamer) partitioned strongly into droplets 5 min after fueling (partition coefficients K>10): SunY K=37 (highest, longest), Hammerhead K=13 (lowest, shortest), consistent with size dependence. As the cycle progressed, RNA appeared increasingly outside droplets until dissolution. - RNA functionality retained: Broccoli aptamer bound DFHBIT ligand at low to moderate Mg2+/K+, with fluorescence increasing ~5-fold in solution; introduction of fuel caused a transient ~30% decrease (turbidity effect) followed by full recovery after fuel consumption. Confocal imaging showed strong co-localization of the Broccoli–DFHBIT complex inside droplets (>10-fold enrichment), indicating preserved folded, active states within droplets.

The results demonstrate a chemically fueled, out-of-equilibrium coacervate system that models key behaviors of membraneless organelles: rapid fuel-responsive emergence, exchange of building blocks, vacuole formation, fragmentation, and controlled dissolution. The reaction cycle spatially separates activation (outside droplets) from accumulation of active product (inside droplets), linking kinetics to droplet composition and stability. Dynamic droplets respond closely to product kinetics, dissolving as product hydrolyzes and releasing RNA; metastable droplets, with higher peptide:RNA ratios and denser packing, exhibit kinetic barriers to disassembly, retaining RNA even after deactivation. The ability to transiently up-concentrate functional RNAs while preserving activity suggests mechanisms by which cells might regulate condensate composition and function via metabolic fuel fluxes, and provides a route for protocell-like behaviors such as transient compartmentalization and division-like fragmentation driven by loss of structural integrity.

This work establishes RNA-containing active coacervate droplets driven by a carbodiimide-fueled peptide activation/deactivation cycle as a model for membraneless organelles and protocells. The system shows fuel-dependent emergence, dynamic evolution (fusion, vacuole formation), kinetically controlled metastability, fragmentation upon decay, and transient up-concentration of functional RNAs that remain active inside droplets. These findings link reaction kinetics to condensate stability and composition, offering a platform to study regulation mechanisms of biological condensates. Future directions include elucidating mechanisms underlying metastability and fragmentation, integrating self-replicating or catalytic RNAs to explore feedback that sustains droplets, and expanding biochemical complexity and environmental conditions to approach cellular relevance.

- Nucleation events could not be directly observed due to rapid onset after fuel addition at the available imaging rates. - Quantification of the anhydride product required indirect HPLC measurement via benzylamine trapping due to anhydride instability and peak overlap, introducing methodological constraints. - Droplet formation and stability were sensitive to ionic strength; turbidity and coacervation decreased with increasing salt, limiting conditions (robust below ~50 mM salt in tested assays). - Experiments were conducted under specific buffer and pH conditions (MES, pH 5.3) and defined temperature ranges, which may limit generalizability to cellular environments. - The kinetic arrest and enhanced stability of metastable droplets are hypothesized to arise from higher peptide:RNA ratios and increased packing, but detailed molecular mechanisms and energy barriers were not fully resolved. - Fragmentation behavior near dissolution was not controlled and only qualitatively characterized; mechanisms remain to be studied.