Liquid-liquid phase separation is thought to underly gene transcription, through the condensation of the large-scale nucleolus, or in smaller assemblies known as transcriptional hubs or condensates. However, phase separation has not yet been directly linked with transcriptional output, and our biophysical understanding of transcription dynamics is poor. Here, we utilize an optogenetic approach to control condensation of key FET-family transcriptional regulators, particularly TAF15. We show that amino acid sequence-dependent phase separation of TAF15 is enhanced significantly due to strong nuclear interactions with the C-terminal domain (CTD) of RNA Pol II. Nascent CTD clusters at primed genomic loci lower the energetic barrier for nucleation of TAF15 condensates, which in turn further recruit RNA Pol II to drive transcriptional output. These results suggest a model in which positive feedback between key transcriptional components drives intermittent dynamics of localized phase separation, to amplify gene expression.Cells organize complex biochemical reactions through compartmentalization into various organelles. Many of these organelles are membrane-less condensates, which form through liquid-liquid phase separation (LLPS), in which biomolecules condense into dynamic liquid droplets [1][2][3][4] . Nuclear condensates are thought to play key roles in regulating the flow of genetic information and include a diverse set of structures ranging from large scale assemblies such as nuclear speckles and the nucleolus, to structures that form on smaller length scales such as Cajal bodies, promyelocytic leukemia (PML) bodies, and gems.The concept that LLPS can drive transcriptional activity has recently received intense attention. The nucleolus is a large nuclear condensate which assembles around actively transcribing ribosomal RNA (rRNA) loci 5,6 , and is thought to facilitate the processing and maturation of rRNA into pre-ribosomal particles. Smaller transcriptional condensates have been proposed to assemble throughout the genome to facilitate transcription of non-ribosomal genes through the RNA polymerase II (Pol II) machinery [7][8][9][10] . These nanoscale transcriptional condensates have been hypothesized to facilitate enhancer-promoter interactions [11][12][13] , possibly through the generation of localized force through surface-tension driven droplet coalescence 14 ..

Forming at the right place and time is important for all structures within living cells. This includes condensates such as the nucleolus, Cajal bodies, and stress granules, which form via liquid-liquid phase separation (LLPS) of biomolecules, particularly proteins enriched in intrinsically-disordered regions (IDRs) 1, 2 (Fig. 1a). In non-living systems, the initial stages of nucleated phase separation arise when thermal fluctuations overcome an energy barrier due to surface tension. This phenomenon can be modeled by classical nucleation theory (CNT), which describes how the rate of droplet nucleation depends on the degree of supersaturation, while the location at which droplets appear is controlled by interfacial heterogeneities 3,4 . In living cells, however, it is unknown whether this framework applies, due to the multicomponent and highly complex nature of the intracellular environment, including the presence of diverse IDRs, whose specificity of biomolecular interactions is unclear [5][6][7][8] . Here we show that despite this complexity, nucleation within living cells occurs through a physical process not unlike that within inanimate materials, but where the efficacy of nucleation sites can be tuned by their biomolecular features. By quantitatively characterizing the nucleation kinetics of endogenous and biomimetic condensates within living cells, we found that key features of condensate nucleation can be quantitatively understood through a CNT-like theoretical framework. Nucleation rates can be significantly enhanced by compatible biomolecular (IDR) seeds, while the kinetics of cellular processes can impact condensate nucleation rates and location specificity. This quantitative framework sheds light on the intracellular nucleation landscape, and paves the way for engineering synthetic condensates precisely positioned in space and time.

Multivalent interactions between deformable mesoscopic units are ubiquitous in biology, where membrane macromolecules mediate the interactions between neighbouring living cells and between cells and solid substrates. Lately, analogous artificial materials have been synthesised by functionalising the outer surface of compliant Brownian units, for example emulsion droplets and lipid vesicles, with selective linkers, in particular short DNA sequences. This development extended the range of applicability of DNA as a selective glue, originally applied to solid nano and colloidal particles. On very deformable lipid vesicles, the coupling between statistical effects of multivalent interactions and mechanical deformation of the membranes gives rise to complex emergent behaviours, as we recently contributed to demonstrate [Parolini et al., Nat. Commun., 2015, 6, 5948]. Several aspects of the complex phenomenology observed in these systems still lack a quantitative experimental characterisation and a fundamental understanding. Here we focus on the DNA-mediated multivalent interactions of a single liposome adhering to a flat supported bilayer. This simplified geometry enables the estimate of the membrane tension induced by the DNA-mediated adhesive forces acting on the liposome. Our experimental investigation is completed by morphological measurements and the characterisation of the DNA-melting transition, probed by in situ Förster Resonant Energy Transfer spectroscopy. Experimental results are compared with the predictions of an analytical theory that couples the deformation of the vesicle to a full description of the statistical mechanics of mobile linkers. With at most one fitting parameter, our theory is capable of semi-quantitatively matching experimental data, confirming the quality of the underlying assumptions.

We present the first direct observations of morphological transitions from macroto micro-phase separation using micrometer-sized asymmetric lipid vesicles exposed to externally added glycolipids (GM1:monosialotetrahexosylganglioside). The transition occurs via an intermediate stripe morphology state. During the transition, monodisperse micro-domains emerge through repeated scission events of the stripe domains. Moreover, we numerically confirmed such transitions using a time-dependent Ginzburg-Landau model, which describes both the intramembrane phase separation and the bending elastic membrane. The experimental and simulation results are in quantitative agreement.