Transcription is spatially organized in the nucleus, concentrating transcription factors, coactivators, and RNA polymerase II into dynamic, membraneless compartments. Such assemblies have been described in many ways, from small complexes of defined stoichiometry to membraneless biomolecular condensates. We argue that these are not competing descriptions but positions along a single continuum, and propose "transcriptional compartment" as an umbrella term for it. We then turn to the functional literature and ask what these compartments have been shown to do, and what they have not. The clearest evidence for a functional role of condensation may emerge in contexts where transcription must generate switch-like responses, such as cell-fate decisions, stress responses, environmental sensing, and disease. Much of the remaining biology, however, sits in the small, transient assemblies between the two extremes. These are the hardest to study, in part because they fall below the diffraction limit of light microscopy and because the interactions that drive their formation usually also carry out their function, so the two cannot be perturbed independently. We survey the experimental and computational toolkit available for this problem and outline the open methodological challenges that remain.
Quail et al. (Thu,) studied this question.