All modern cells, whether prokaryotic or eukaryotic, unicellular or multicellular, have one thing in common – the components of the cell are separated from the environment, or, in other words, they are compartmentalised. This feature is one of the foundations of living systems and confining biomolecules in compartments was likely one of the key steps in the emergence of life.
When trying to reconstruct the first cell, compartments made by lipid vesicles may seem to be a natural way to go, as modern cells are surrounded by lipid bilayers. However, their lack in permeability to many polar life building blocks (like nucleic acids or peptides) brings into question their suitability as the simplest protocellular reaction compartments. Several alternative ways to compartmentalise life’s building blocks have been suggested, and coacervates – water‑rich droplets formed upon liquid‑liquid phase separation (LLPS) – are considered a strong candidate, free of limitations characteristic for lipid membranes, and capable
Coacervates are already present in biological world as membraneless organelles, liquid‑like assemblies of mostly proteins and nucleotides. Usually, their components are relatively complex molecules – too large and complex for prebiotic systems. So, we set ourselves a goal to find a minimal molecular motif which can form coacervates. In our recent paper in Nature Chemistry, we show that coacervates can be formed by combining in single molecule two hydrophobic dipeptide fragments and a hydrophilic linker  The design was inspired by the sticker‑and‑spacer motif that was recently described to be an important characteristic of phase‑separating proteins 
Proteins that undergo LLPS often contain aromatic amino acid residues grouped in fragments called stickers, separated by flexible hydrophilic regions called spacers. So far, the sticker‑and‑spacer motif has been described to drive LLPS of only relatively large proteins. We hypothesised that linking two hydrophobic dipeptides with polar spacer may result in the formation of liquid droplets rather than irreversible solid aggregates, which are often formed by simple aromatic peptides. Our first choice for the spacer was a cystamine moiety (due to its flexibility, hydrophilicity and possibility to cleave it with reducing agents) and for the sticker – an l‑phenylalanyl‑l‑phenylalanine moiety (FF), known for its propensity to from fibre‑like aggregates. In short, we called the conjugate FFssFF.
FFssFF dissolved in water formed a solution at pH below 6 and concentrations up to 15 mg/ml. However, when the pH is increased above 7, the solution of FFssFF became turbid. Under these conditions FFssFF formed condensed droplets with a typical size of 1 to 10 μm. The droplets were able to fuse, spread, deform and ultimately separate into a bulk phase. While these droplets contained very high peptide concentration (estimated to be 1000 times higher than in the surrounding diluted phase), they were still rich in water (75 ± 10 % (w/w) at pH 8), which makes them substantially different from oil‑in‑water droplets.
We decided to explore the general nature of the sticker‑and‑spacer design for small coacervate‑forming molecules. We synthesised a series of similar derivatives with different hydrophobic dipeptides and different linkers.
Coacervate droplets as microreactors
Finding a simple molecule that undergoes LLPS doesn’t yet prove that coacervates could have served as primeval compartments capable of hosting chemical reactions. To show that droplets formed by the small peptide derivatives are attractive candidates for protocells, we wanted to prove their ability to work as microreactors. First, we investigated their ability to accumulate various molecules and increase their local concentration. Droplets formed by FFssFF could sequester various aromatic dyes, but also ssDNA and RNA. What is more, by incubating a short RNA hairpin with the FFssFF coacervates we have shown that the droplets can not only accumulate nucleic acid duplex, but also promote their melting, as evidenced by an increase in fluorescence intensity – and this suggests that the coacervate environment is distinct from the surrounding solution.
As we have observed that the FFssFF coacervates increase local concentration of various molecules and provide a distinct environment from the surrounding diluted phase, we set out to investigate how they could affect kinetics of anabolic reactions. We selected aldol condensation and hydrazone formation as model reactions, which are both slow at neutral pH and without catalyst. These reactions occurred up to 340 times faster in the presence of coacervate droplets, depending on the type of stickers and spacer. Interestingly, higher local concentration (which we could estimate by knowing the partition coefficients) of substrates inside the droplets can’t fully account for the reaction rates enhancement. Therefore, we concluded that the acceleration is caused by a combination of increased concentration and lowering of the reaction energy barrier (due to the apolar environment inside the droplets).
Did life begin in coacervate droplets? We are not able to say it yet, and maybe we will never be. However, these small, coacervate‑forming molecules may provide a stepping stone for the development of a wide range of new protocells with catalytic properties. Our results show one of the first examples of small‑molecule anabolic reactions in droplets, and we hope that the field of chemically active coacervates  will attract more attention in the near future. What’s more, they should provide better understanding of the molecular principles underlying liquid‑liquid phase separation.
- Abbas, M., Lipiński, W. P., Nakashima, K. K., Huck, W. T. S. & Spruijt, E. A Short Peptide Synthon for Liquid-Liquid Phase Separation. Nat. Chem. (2020).
- Martin, E. W. et al. Valence and patterning of aromatic residues determine the phase behavior of prion-like domains. Science 367, 694–699 (2020).
- Donau, C. et al. Active coacervate droplets as a model for membraneless organelles and protocells. Nat. Commun. 11, 1–10 (2020).