Endosomal membrane budding patterns in plants

Abstract

Multivesicular endosomes (MVEs) sequester membrane proteins destined for degradation within intralumenal vesicles (ILVs), a process mediated by the membrane-remodeling action of Endosomal Sorting Complex Required for Transport (ESCRT) proteins. In Arabidopsis, endosomal membrane constriction and scission are uncoupled, resulting in the formation of extensive concatenated ILV networks and enhancing cargo sequestration efficiency. Here, we used a combination of electron tomography, computer simulations, and mathematical modeling to address the questions of when concatenated ILV networks evolved in plants and what drives their formation. Through morphometric analyses of tomographic reconstructions of endosomes across yeast, algae, and various land plants, we have found that ILV concatenation is widespread within plant species, but only prevalent in seed plants, especially in flowering plants. Multiple budding sites that require the formation of pores in the limiting membrane were only identified in hornworts and seed plants, suggesting that this mechanism has evolved independently in both plant lineages. To identify the conditions under which these multiple budding sites can arise, we used particle-based molecular dynamics simulations and found that changes in ESCRT filament properties, such as filament curvature and membrane binding energy, can generate the membrane shapes observed in multiple budding sites. To understand the relationship between membrane budding activity and ILV network topology, we performed computational simulations and identified a set of membrane remodeling parameters that can recapitulate our tomographic datasets.

Keywords: ESCRT; multivesicular endosomes; vesiculation.

Fig. 1.

MVEs in land plants. (A) Tomographic reconstructions from representatives of the major clades of land plants as well as budding yeast and Chara. (B) Quantitative analysis of MVE diameter across species. (C) Quantitative analysis of ILV diameter. Most species show ILVs that fall into two types based on their sizes and content electron-density. Asterisks indicate the population of ILVs of larger size and translucent lumen. (D) Example of two types of ILVs in M. polymorpha; asterisks indicate large, electron-translucent ILVs whereas red arrows point at smaller, electron-dense ILVs. (E) Percentage of concatenated, free, and single ILV buds across species. Different letters on graphs indicate significant difference (P < 0.05) calculated by a chi-squared test. (Scale bars, 50 mm).

Fig. 2.

Representative examples of ILV analysis in MVEs of Arabidopsis root cells. (A) Tomographic slice and (B) tomographic reconstruction of an MVE with discrete ILV networks and free ILVs depicted in different colors. (C) Free ILVs, single buds, and concatenated ILV networks from MVE shown in (A) and (B), were analyzed separately and depicted as 2D simplified topologies, with ILVs represented by circles and membranous bridges, by lines between ILVs. Asterisks indicate bridges still connected to the limiting membrane (budding site) and the number inside each ILV reflects the number of membranous bridges connected to it. Parentheticals indicate the number of times a given topology was found in the reconstructed MVE. (D) Graph of average number of membranous bridges per ILV in the species under study. Different letters on graphs indicate significant difference (P < 0.05) calculated by one-way ANOVA followed by Tukey’s test. (E) Graph depicting the average number of bridges per ILV against the average number of ILVs in budding sites (multiplicity). Regression analysis between the two variables retrieved an R² value of 0.94. The color of each circle represents the average ILV diameter for that species.

Fig. 3.

Simulation of changes in ESCRT filament geometries and associated membrane deformation during formation of concatenated ILV networks. (A) ESCRT-III filaments of radius R consist of interconnected three-beaded subunits with the angle α between subunits (Left). The membrane-binding faces are depicted in blue and the strength of the adhesion is controlled by ϵ_ema (Right). The angle τ is the angle established between the radial axis and the triplet subunit axis. A τ value larger than 0º defines the tilted state of the filament. (B) During the simulation, the filament was allowed to break into 2 or more independent filaments. (C) Snapshots of the membrane and ESCRT-III filament intermediates during the formation of concatenated ILVs following different possible pathways. The changes of the filament properties (adhesion energy ϵ_ema, the α angle between subunits or the filament radius, and the tilt angle τ) during the simulation are indicated between intermediate simulation steps. In most cases, properties changed equally for all filaments. Only during the transition between the f and m states, the radius of the left and right filaments R(l + r) decreased and the tilt angle of the left and right filaments τ(l + r) increased, while the filament properties of the lower filament remained constant. Similarly, during the transition between the states f' and k, the tilt angle of the left filament τ(l) decreased, while the adhesion energy of all three filaments increased. The combination of parameter changes resulted in single buds (h), three types of linear ILV chains (g, j, k), free ILV networks (i), and two types of double budding sites (l and m). Also Movie S1. (D) Formation of a pore in a simulated double budding site. (E) Continuum model for pore formation in double budding sites. Condition for membrane pore formation due to ESCRT filament adhesion according to Eq. 1. The solid yellow line marks the minimal adhesion energy density of ESCRT filaments μ as a function of the preferred ESCRT radius R_E so that membrane pore formation becomes energetically favorable. In the blue region, membrane pore formation is energetically favorable while in the red region membrane pore formation is energetically unfavorable.

Fig. 4.

Simulation of ILV network formation. (A) Diagrams depicting the three possible actions allowed to occur in the simulation. Asterisks indicate bridges still connected to the limiting membrane (budding site) and the letter and number inside each ILV reflects the order in which they were generated. (B) Distribution of the 1,000 repeats of 63 simulations each using the parameters p1 = 0.54, p2 = 0.01, p3 = 0.45. The red arrows indicate the values corresponding to the statistical analysis performed on the 63 ILV networks from the pine MVEs.

Fig. 5.

Representative examples of MVEs of Arabidopsis mutants lacking ESCRT components. (A) Tomographic reconstruction of mutant MVEs and 2D ILV topologies. Asterisks indicate bridges still connected to the limiting membrane (budding site) and the number inside each ILV reflects the number of membranous bridges connected to it. Parentheticals indicate the number of times a given topology was found in the reconstructed MVE. (B) Percentage of concatenated, free, and single ILV buds across mutant MVEs. Different letters on graphs indicate significant difference (P < 0.05) calculated by a chi-squared test. (C) Percentage of budding sites with 1 to 4 interconnected ILV buds in wild type (WT) and mutant MVEs. (D) Percentage of ILVs associated with 0 to 5 membranous bridges. (E) Graph depicting the average number of bridges per ILV against the average number of ILVs in budding sites (multiplicity) in WT and mutant MVEs. (Scale bars, 50 mm).