Evolution: On a bender--BARs, ESCRTs, COPs, and finally getting your coat

Abstract

Tremendous variety in form and function is displayed among the intracellular membrane systems of different eukaryotes. Until recently, few clues existed as to how these internal membrane systems had originated and diversified. However, proteomic, structural, and comparative genomics studies together have revealed extensive similarities among many of the protein complexes used in controlling the morphology and trafficking of intracellular membranes. These new insights have had a profound impact on our understanding of the evolutionary origins of the internal architecture of the eukaryotic cell.

Figure 1.

The endomembrane system: establishment, elaboration, and sculpting across evolutionary time. Top: cellular configuration of intracellular membrane architectural features. Second from top: molecular machineries that are associated with the endomembrane system, and predicted points of origin in eukaryotic evolution. Third from top: diagrams of cellular architectures to illustrate the origins of phagocytosis, internal membranes, and endosymbiotic organelles, and how these relate to the origins of cellular systems and the first (FECA) and last (LECA) eukaryotic common ancestors. The suggested sequence of events, although being the one that we favor, is not the only possible order; it is still unresolved at which points the nucleus, flagellum, mitochondria, phagocytosis, and endocytosis developed. Bottom: category of cell, using a generalized terminology.

Figure 2.

Association of membrane-deforming machineries with the endomembrane system. An overview of a generic eukaryotic endomembrane system, with major compartments, endosomes, and exocytic/endocytic pathways, is shown. The diagram is highly schematic and is used to represent the major organelles, exocytic and endocytic pathways present in a great many eukaryotes—specific lineage-specific features are omitted for clarity. Arrows show known and presumed major trafficking routes and their directions—again the representation is generic and not all routes operate in all cells. Membranes are colored to illustrate the participation of major protocoatomer (red), ESCRT (green), and BAR (blue) domain-associated complexes.

Figure 3.

Architectures of membrane deformation complexes. Depicted are, from top, atomic level structures, schematics for the molecular architectures of subunit assemblies, architecture of the coats lattice, and depictions of how these complexes may assemble at the membrane. In all cases, α-structures are shown in pink and β-structures in blue. For complex assemblies at the membrane, the deformation subunits are shown in red and blue, and associated factors responsible for cargo recognition or other interactions are in gray. Note that the topology of membrane deformation between protocoatomer and most BAR domain complexes is 180° out of phase with the ESCRT complex. The figure incorporates ribbon diagrams for the crystallographic structures of the Nup145c/Sec13 dimer (PDB: 3jro; Brohawn and Schwartz, 2009); Sec13/31 dimer (PDB: 2pm6 and 2pm9; Fath et al., 2007); the N-terminal β-propeller domain of human Nup133 (PDB: 1xks; 75–482; Berke et al., 2004) and the human C-terminal helical domain (PDB: 3i4r; 517–1156; Whittle and Schwartz, 2009); two BAR domain homodimers, human formin-binding protein 17 (PDB: 2efl; 1–300; Shimada et al., 2007) and human PACSIN2/Syndapin II (PDB: 3abh; 16–304); the tetrameric ESCRT III complex (PDB: 2zme) consisting of vacuolar-sorting protein SNF8 (34–250), two copies of vps36 (172–386), and one copy of vps25 (4–103); the clathrin heavy chain, based on the structure of a bovine C-terminal helical region (PDB: 3lvh; 1077–1630; Wilbur et al., 2010) and the rat N-terminal region (PDB: 1bpo; 1–494; ter Haar et al., 1998); and finally a portion of the yeast COPI coat (PDB: 3mkq; β subunits 1–814; β subunits 624–818; Lee and Goldberg, 2010). Note the reverse topology of BAR domain complexes is omitted for clarity.