Many membraneless organelles are liquid-like domains that form inside the active, viscoelastic environment of living cells through phase separation. To investigate the potential coupling of phase separation with the cytoskeleton, we quantify the structural correlations of membraneless organelles (stress granules) and cytoskeletal filaments (microtubules) in a human-derived epithelial cell line. We find that microtubule networks are substantially denser in the vicinity of stress granules. When microtubules are depolymerized, the sub-units localize near the surface of the stress granules. We interpret these data using a thermodynamic model of partitioning of particles to the surface and bulk of the droplets. In this framework, our data are consistent with a weak (≲ k B T ) affinity of the microtubule sub-units for stress granule interfaces. As microtubules polymerize, their interfacial affinity increases, providing sufficient adhesion to deform droplets and/or the network. Our work suggests that proteins and other objects in the cell have a non-specific affinity for droplet interfaces that increases with the contact area and becomes most apparent when they have no preference for the interior of a droplet over the rest of the cytoplasm. We validate this basic physical phenomenon in vitro through the interaction of a simple protein–RNA condensate with microtubules. Many organelles in the cell are not encapsulated in a membrane—they are liquid-like domains formed through phase separation. The liquid-like nature of such domains leads to adhesive interactions between the cytoskeleton filaments and organelles.

[This corrects the article DOI: 10.1038/s41567-022-01537-8.].

Membrane-less organelles (MLOs) are liquid-like subcellular compartments that form through phase separation of proteins and RNA. While their biophysical properties are increasingly understood, their regulation and the consequences of perturbed MLO states for cell physiology are less clear. To study the regulatory networks, we targeted 1,354 human genes and screened for morphological changes of nucleoli, Cajal bodies, splicing speckles, PML nuclear bodies (PML-NBs), cytoplasmic processing bodies, and stress granules. By multivariate analysis of MLO features we identified hundreds of genes that control MLO homeostasis. We discovered regulatory crosstalk between MLOs, and mapped hierarchical interactions between aberrant MLO states and cellular properties. We provide evidence that perturbation of pre-mRNA splicing results in stress granule formation and reveal that PML-NB abundance influences DNA replication rates and that PML-NBs are in turn controlled by HIP kinases. Together, our comprehensive dataset is an unprecedented resource for deciphering the regulation and biological functions of MLOs.

The lipid content of the plasma membrane is dynamically regulated to maintain cellular homeostasis, but the molecular links between membrane stress and sphingolipid synthesis have remained elusive. Walther, Loewith and colleagues report that membrane stretching causes redistribution of Slm proteins, which then promote sphingolipid synthesis through activation of the TORC2–Ypk signalling pathway.

We used chicken retinospheroids (RS) to study the nuclear architecture of vertebrate cells in a three-dimensional (3D) cell culture system. The results showed that the different neuronal cell types of RS displayed an extreme form of radial nuclear organization. Chromatin was arranged into distinct radial zones which became already visible after DAPI staining. The distinct zones were enriched in different chromatin modifications and in different types of chromosomes. Active isoforms of RNA polymerase II were depleted in the outermost zone. Also chromocenters and nucleoli were radially aligned in the nuclear interior. The splicing factor SC35 was enriched at the central zone and did not show the typical speckled pattern of distribution. Evaluation of neuronal and non-neuronal chicken tissues showed that the highly ordered form of radial nuclear organization was also present in neuronal chicken tissues. Furthermore, the data revealed that the neuron-specific nuclear organization was remodeled when cells spread on a flat substrate. Monolayer cultures of a chicken cell line did not show this extreme form of radial organization. Rather, such monolayer cultures displayed features of nuclear organization which have been described before for many different types of monolayer cells. The finding that an extreme form radial nuclear organization, which has not been described before, is present in RS and tissues, but not in cells spread on a flat substrate, suggests that it would be important to complement studies on nuclear architecture performed with monolayer cells by studies on 3D cell culture systems and tissues.

The conserved target of rapamycin (TOR) kinase regulates cell growth and homeostasis and exists in two functionally distinct complexes, TORC1 and TORC2. The authors explored the spatial organization of the two TOR complexes in yeast and found them in distinct places: TORC2 localizes to the plasma membrane, whereas TORC1 localizes to the vacuolar membrane. TORC2 exists in small dynamic foci akin to signaling nanoclusters like those observed for Ras proteins. These clusters occupy a distinct membrane compartment. Avo1, a subunit of TORC2, binds the plasma membrane lipid PI (4, 5) P2 via a PH-like domain that is required for TORC2 function and localization. However, this region of the protein can be replaced by a short heterologous sequence tag that directs protein lipidation and targeting to the plasma membrane without loss of function, arguing that plasma membrane localization of TORC2 is …

The conserved target of rapamycin (TOR) kinases regulate many aspects of cellular physiology. They exist in two distinct complexes, termed TOR complex 1 (TORC1) and TOR complex 2 (TORC2), that posses both overlapping and distinct components. TORC1 and TORC2 respond differently to the drug rapamycin and have different cellular functions: whereas the rapamycin-sensitive TORC1 controls many aspects of cell growth and has been characterized in great detail, the TOR complex 2 is less understood and regulates actin polymerization, cell polarity, and ceramide metabolism. How signaling specificity and discrimination between different input signals for the two kinase complexes is achieved is not understood. Here, we show that TORC1 and TORC2 have different localizations in Saccharomyces cerevisiae. TORC1 is localized exclusively to the vacuolar membrane, whereas TORC2 is localized dynamically in a previously unrecognized plasma membrane domain, which we term membrane compartment containing TORC2 (MCT). We find that plasma membrane localization of TORC2 is essential for viability and mediated by lipid binding of the C-terminal domain of the Avo1 subunit. From these data, we suggest that the TOR complexes are spatially separated to determine downstream signaling specificity and their responsiveness to different inputs.