Liquid-liquid phase separated (LLPS) states are key to compartmentalise components in the absence of membranes, however it is unclear whether LLPS condensates are actively and specifically organized in the sub-cellular space and by which mechanisms. Here, we address this question by focusing on the ParAB S DNA segregation system, composed of a centromeric-like sequence (parS), a DNA-binding protein (ParB) and a motor (ParA). We show that parS-ParB associate to form nanometer-sized, round condensates. ParB molecules diffuse rapidly within the nucleoid volume, but display confined motions when trapped inside ParB condensates. Single ParB molecules are able to rapidly diffuse between different condensates, and nucleation is strongly favoured by parS. Notably, the ParA motor is required to prevent the fusion of ParB condensates. These results describe a novel active mechanism that splits, segregates and localises non-canonical LLPS condensates in the sub-cellular space.

We study the 'Large Volume Scenario' on explicit, new, compact, four-modulus Calabi-Yau manifolds. We pay special attention to the chirality problem pointed out by Blumenhagen, Moster and Plauschinn. Namely, we thoroughly analyze the possibility of generating neutral, non-perturbative superpotentials from Euclidean D3-branes in the presence of chirally intersecting D7-branes. We find that taking proper account of the Freed-Witten anomaly on non-spin cycles and of the Kähler cone conditions imposes severe constraints on the models. Nevertheless, we are able to create setups where the constraints are solved, and up to three moduli are stabilized. C. Fourth model: A matterless model 35 C.1 The resolved P 4 1,1,1,3,3 (9) /Z 3 : 0 0 0 2 1 geometry 35 C.2 Moduli stabilization 37 IntroductionThe 'Large Volume Scenario' (LVS), developed in [1] is a new strategy for stabilizing the Kähler moduli in IIB Calabi-Yau orientifold compactifications. This strategy can be seen as a cousin of the KKLT strategy [2]. In both cases, one first stabilizes the axio-dilaton and complex structure moduli by means of the flux induced Gukov-Vafa-Witten superpotential, and then one tries to stabilize the Kähler moduli by non-perturbative effects such as E3branes (Euclidean D3-branes), and gaugino condensation. The key difference between these two strategies lies in the fact that the LVS admits non-supersymmetric anti-de Sitter minima, whereby the Calabi-Yau volume is exponentially large w.r.t. the size of the E3brane, and, at fixed g s , it is independent of the flux superpotential W 0 . This latter fact implies that this non-perturbative stabilization of the Kähler moduli will not mess up the complex structure stabilization. Other advantages of this scenario are explained in [3].The key requirement to construct an LVS model, is to find a Calabi-Yau threefold with h 2,1 > h 1,1 > 1, and such that the volume of the manifold is driven by the volume of a single 'large' four-cycle, and that the rest of the four-cycles contribute negatively to the overall volume. This structure has been dubbed the 'Swiss cheese' structure. Because it is possible to make cycles small while keeping the CY large, we can have E3-instantons that make large contributions and have a large volume vacuum. These instanton effects now becoming important, actually compete against α ′ corrections to the Kähler potential. Having these 'small', shrinkable cycles also serves another useful purpose. If one places MSSM-like stacks of D7-branes on them, by going to this large volume limit where these are made small, one effectively decouples the gauge theory on the brane from the UV dynamics encoded by the rest of the Calabi-Yau data. In this way, one addresses the comment in [4], which points out a drawback of generic models: Namely, that making the volume of the CY large will typically force one to scale up the cycles on which branes are wrapped.In [5], Blumenhagen et al have shown that the standard two-step model building paradigm, where one first stabilizes the closed string mod...

We systematically construct a large number of compact Calabi-Yau fourfolds which are suitable for F-theory model building. These elliptically fibered Calabi-Yaus are complete intersections of two hypersurfaces in a six dimensional ambient space. We first construct three-dimensional base manifolds that are hypersurfaces in a toric ambient space. We search for divisors which can support an F-theory GUT. The fourfolds are obtained as elliptic fibrations over these base manifolds. We find that elementary conditions which are motivated by F-theory GUTs lead to strong constraints on the geometry, which significantly reduce the number of suitable models. The complete database of models is available at [1]. We work out several examples in more detail.