Biomolecular condensates, the membrane-less droplets that cells use to organize chemistry without enclosing it, form by liquid–liquid phase separation, LLPS, driven by multivalent weak interactions among proteins and RNA. Replicating that logic with short synthetic peptides inside living cells is conceptually attractive but practically demanding: reaching the critical concentration for LLPS typically requires local peptide levels an order of magnitude higher than those needed to seed solid aggregates, and the resulting droplets are sensitive to salt, pH, and temperature. Prior work on intracellular self-assembly had produced nanofibers and hydrogels, but crossing from liquid-to-solid into liquid-to-liquid phase transitions with minimalist synthetic inputs had not been achieved. Solving that problem would provide a platform for reversible, stimulus-responsive compartments that could be switched on and off inside living cells without permanent structural commitment.
Researchers working with Tanja Weil and David Ng in the Department of Synthesis of Macromolecules at the Max Planck Institute for Polymer Research, published in J. Am. Chem. Soc., reasoned that rather than fighting the intracellular environment, a well-designed peptide could exploit it. They synthesized Fmoc-K(TPP)RGRGR-CONH2, a seven-residue cationic peptide carrying an N-terminal fluorenylmethoxycarbonyl block, three arginine residues alternating with glycines in a pattern that mimics RG motifs of natural RNA-binding proteins, and a triphenylphosphonium, TPP, side chain originally intended to direct the peptide to mitochondria. In vitro calorimetry revealed that polymeric RNA binds the peptide with a dissociation constant of 3.5 ± 0.3 μM, more than two orders of magnitude tighter than ATP, and drives coacervation at peptide concentrations 30-fold lower than ATP requires. Inside A549 human lung adenocarcinoma cells, the arginine content overrides the TPP targeting signal and steers the peptide toward the RNA-dense nucleolus, where complex coacervates nucleate within one minute. The Fmoc π-block, the number of Gly–Arg repeats, and the TPP side chain each proved essential: due to a decrease in RNA affinity as evidenced by in vitro experiments.
The most consequential feature of the system is its reversibility. Washing out the peptide triggers droplet dissolution within about sixty minutes and allows cellular viability to recover to nearly 80%, in contrast to the irreversible fiber-forming peptides that dominate the intracellular self-assembly literature. This substrate-dependent on/off behavior provides a conceptual foundation for reversible compartment engineering inside living cells, transient metabolic modulation, and the design of synthetic reaction spaces that operate in dynamic exchange with the cellular environment. The full structural requirements, kinetic data, and cell-recovery experiments are reported in the original publication.
