Figure 1.
Lateral compartmentalization, assessed by Fluorescence Loss In Photobleaching (FLIP) assays, reflects the diffusion barriers' strength.
The yeast nucleus buds into the daughter cell in early stages of anaphase, elongating into a dumbbell shape in late anaphase. Diffusion barriers have been estimated to locate somewhere between the mother and daughter nuclear halves. Compartmentalization is measured by continuously bleaching the mother lobe while simultaneously measuring the fluorescence decay over time in mother and daughter lobes, separately.
Figure 2.
Sphingolipids self-organize into tightly packed, rigid and thicker domains within membranes.
The increased viscosity within these domains causes membrane-bound proteins to diffuse at a lower rate. Moreover, the phase change at the boundary between the domain and the rest of the membrane may work against proteins trying to diffuse into the domain. We call this effect the protein exclusion effect. Measures indicated are: , diameter of membrane inclusions;
, thickness of diffusive media (membranes and periplasm);
, bulk viscosities (measured in [Pa][s]);
, surface viscosities (measured in [Pa][s][m]); Pin and Pout, probability of proteins diffusing into and out of the sphingolipid domain, respectively.
Figure 3.
Possible diffusion barrier scenarios, constituted by different specialised lipid domain configurations.
For early anaphase: (A) a ring-shaped domain at the neck of mitotic nuclei in early anaphase. For late anaphase: (B) a single ring-shaped domain centred at the bridge between nuclear lobes, (C) a set of parallel rings uniformly spaced along the bridge, and (D) a homogeneously distributed domain spanning the entire bridge length.
Table 1.
Estimated Pin values for membrane proteins in different specialised lipid domain configurations.
Figure 4.
Barrier permeabilities of different specialised lipid domain configurations and for maker proteins at the ONM and INM.
The transmission coefficient of the barrier depends directly on the partition coefficient and the diffusion rate within the domain, and is inversely dependent on its thickness (see Methods). The error bars are associated to the heterogeneous distribution of surface areas (21 nuclei in EA and 34 in LA). For the single ring configuration in LA, a ring of 100 nm wide was assumed for diffusion of NPCs (asterisks). For the NPC, we considered both scenarios where the domain lies only at the ONM or at both ONM and INM. Effective diffusion coefficients were fixed as in Table S2 and we fixed Pout = 100% in all cases. The corresponding Pin values are shown in Table 1.
Figure 5.
Comparing diffusion barrier scenarios constituted by different specialised lipid domain configurations in LA.
°CP (green) and °CP−1 (orange) ratios of Nsg1-GFP and GFP-Src1 plotted against the position of the bleaching spot relative to the bridge length (starting at the junction between the mother lobe and the bridge, ending where the latter joins the daughter lobe). Pin values were fixed as in Table 1. Experimental data (black) reconstructed from Fig. 4C in [2].
Figure 6.
Estimated number n of proteins at the ring constraining lateral diffusion in EA nuclei.
Average deviations (in percentages) of stochastic simulations from the experimental mean for each experimental time step. Mother and daughter lobe deviations are calculated as the absolute value of the difference between simulations and FLIP experiments. The average deviation is the mean of both nuclear lobes' deviations over time. The best fit (i.e. minimal deviation) is shown as a continuous line. For NPC data, we assumed the specialised lipid domain lies at both the ONM and INM, whereas the protein ring is present at the ONM only. Effective diffusion coefficients were fixed as in Table S2 and we fixed Pout = 100% in all cases. Pin values were fixed as in the homogeneous domain scenario in LA (Table 1). For Nsg1-GFP, ns stands for the number of septin proteins in a double ring arrangement.