Introduction: a new grammar for cellular organization

For much of the twentieth century cell biologists thought of organization as a story of membranes: discrete organelles boxed off by lipid bilayers. Yet the nucleus, that dense hub of DNA, contains structures without membranes — nucleoli, speckles, Cajal bodies — that rise and fall with the cell’s needs. The last decade has revealed these not as static curiosities but as dynamic, physics-driven assemblages: biomolecular condensates formed by phase separation. The insight reframes how genes are regulated and how disorder in molecular assemblies can cause disease.

A short history: from nucleoli to liquid droplets

The nucleolus was observed by microscopists in the nineteenth century as a conspicuous dark body within the nucleus. Through the twentieth century researchers cataloged many such bodies and debated whether they were functional compartments or mere aggregates. The conceptual leap came from physicists and cell biologists who asked whether biology might use the same principles that make oil separate from water.

In 2009 Cliff Brangwynne and Tony Hyman provided a decisive demonstration: P granules in C. elegans behave like liquid droplets, fusing, dripping, and exchanging molecules with their surroundings. Those experiments — visual, quantitative and elegant — brought liquid–liquid phase separation (LLPS) into cell biology. From that spark, a constellation of researchers including Michael Rosen, Steve McKnight, Jeannie Banani, and others built the molecular grammar: intrinsically disordered regions and multivalent interactions drive reversible condensation.

Key figures and experiments

  • Cliff Brangwynne and Tony Hyman — showed P granules are liquid-like and proposed phase separation as a mechanism for membraneless compartments.
  • Michael Rosen and Steve McKnight — elucidated how low-complexity domains and multivalent protein interactions produce hydrogel-like or liquid phases in vitro, linking sequence features to material behavior.
  • Jeannie Banani, Clifford Brangwynne, and others — synthesized the concept of biomolecular condensates as cellular organizers in influential reviews and experiments.
  • Richard Young, Job Dekker, and colleagues — connected condensates to transcriptional control, showing that transcription factors, Mediator and RNA polymerase II can be concentrated into enhancer-associated hubs that behave like condensates.

Why it was revolutionary

Phase separation introduced a physical mechanism that complements the biochemical view of regulation. Instead of relying only on stable complexes or membrane boundaries, cells can create reversible concentration gradients, speed up reactions by locally concentrating enzymes and substrates, and build responsive compartments that form and dissolve in minutes. This model explains phenomena ranging from the self-assembly of the nucleolus to the punctate foci of signaling proteins and the transient bursts of transcription at enhancers.

Condensates as gene‑regulation hubs

At the heart of gene regulation is a problem of numbers: transcription factors and coactivators are often present at low average concentrations but must find and activate specific genes. Condensates solve this by raising local concentrations at particular genomic loci — enhancers, promoters, or clusters of regulatory elements — thereby increasing the probability of productive interactions among Mediator, transcription factors, and RNA polymerase II. Experimental work shows that activation domains of transcription factors and the carboxy-terminal domain (CTD) of RNA Pol II can engage in phase separation; post-translational modifications such as phosphorylation tune the dynamics, allowing condensates to recruit or release polymerase at the right time.

Implications for disease and technology

The same physics that makes condensates useful can also make them dangerous. Proteins involved in neurodegeneration — FUS, TDP‑43, hnRNPA1 — contain low-complexity domains that form condensates; mutations or cellular stress can drive a liquid-to-solid transition, creating persistent aggregates associated with ALS and frontotemporal dementia. Understanding the rules of condensation opens therapeutic possibilities: small molecules or chaperones that preserve fluidity, or engineered domains that redirect condensate composition.

Where we still grapple

Many questions remain. How do specific DNA sequences and chromatin states nucleate condensates at particular loci? What are the precise material states inside living nuclei — are they purely liquid, viscoelastic gels, or something in-between? And crucially, how do cells ensure specificity so that condensates do not trap the wrong molecules? Addressing these questions requires quantitative live-cell imaging, biophysical measurements, and genetics in parallel.

Conclusion: a living physics

Phase separation gave biology a new verb: to condense. It has transformed our thinking about nuclear organization from static maps to dynamic landscapes that cells sculpt on demand. The discovery united physicists and cell biologists, tied molecular sequence to material property, and offered concrete explanations for both vitality and vulnerability in the nucleus. As we learn to read and write the molecular grammar of condensates, we gain new levers to understand development, disease and to engineer cells that organize chemistry with the elegance of liquid droplets.