@misc{oai:ir.soken.ac.jp:00004063, author = {中村, 健介 and ナカムラ, ケンスケ and NAKAMURA, Kensuke}, month = {2016-02-26}, note = {In order to gain insights into regulations of various membrane structures in cell, structural and functional studies were carried out on two target proteins: Arfaptin, a protein involved in tubule formation at trans-Golgi network; and Atg16L, a protein involved in formation of autophagosomes. The BAR (Bin/Amphiphysin/Rvs) domain forms a curved helical homodimer that can sense or induce the curvature of the membrane it associate with. BAR domain binds to acidic membranes with its basic concaved face in a rather nonspecific manner. This is in contrast to other membrane-binding domains, many of which recognize the head groups of specific phospholipids. BAR domains are often found in conjunction with other membrane-binding modules, such as a PH (pleckstrin homology) domain and a PX (phox homology) domain, which define spatial and functional specificities in membrane association. For instance, many members of the Sorting Nexin (SNX) family, including SNX9, have two membrane-binding properties attributable to its curvature sensing BAR domain and phosphoinositide-binding PX domain; this allows for the selective targeting to high curvature sub-domains of endosomal compartments enriched in specific phosphopnositides. Upon associating with target membranes, the SNX PX-BAR domain drives membrane tubulation for tubule-based sorting. BAR domain-containing protein Arfaptin functions in both tubule formation and stabilization; Arfaptin-1 stabilizes the fission sites of secretory granules at the trans-Golgi network; Arfaptin-2 induces the membrane tubules in Golgi-membranes. The two isoforms have amino acid sequence similarity of 68% and share similar domain structures. Both isoforms localize in the Golgi region, when exogenously expressed in cells, through the association with small GTPase Arf-like 1 (Arl1). Arfaptin-2 colocalizes with Arl1 on dynamic vesicular and tubular structures emanating from the Golgi, suggesting that Arl1 regulates Arfaptin-mediated membrane deformation at the trans-Golgi. The two isoforms of Arfaptin are similarly affected by Arl1, but they associate with different regulators. Arfaptin-2, but not Arfaptin-1, associates with another small GTPase Rac1. The crystal structures of the Rac1–Arfaptin-2 BAR domain complex, reported by Tarricone, et al, show how Rac1 binds to the concaved face of the Arfaptin-2 BAR homodimer, in the way that would interfere with the membrane-association of the BAR domain. In addition, Rac1 also interferes with the Arf-association of Arfaptin-2. Hence, the function of Arfaptin-2 is intricately regulated by the crosstalk between the small GTPases, Arf/Arl1 and Rac1. However, the molecular basis of this crosstalk remained unrevealed, for the complex structure of Arl1–Arfaptin-2 was not known. The crystal structures were determined for Arfaptin-1 in free form and Arfaptin-1 or Arfaptin-2 in complex with Arl1. In the complex structures, two molecules of Arl1 were symmetrically bound on each side of the crescent-shaped homodimer of Arfaptin BAR domain, leaving the concave face open for membrane association. This conformation provided the structural basis for recruitment of Arfaptins onto Golgi membranes by Arl1. Arl1 and Rac1, another binding partner of Arfaptin-2, bound to Arfaptin-2 with the mM-order dissociation constants in Surface Plasmon Resonance (SPR) experiments. Structural comparison between Rac1–Arfaptin-2 and Arl1–Arfaptin-2, combined with SPR experiments, indicated that Rac1 interferes with the one of two molecules of Arl1 on Arfaptin-2. These results provided structural basis for the recruitment of Arfaptins onto Golgi membranes by Arl1 and the crosstalk between Arf/Arl1 and Rac1. Autophagy is the basic catabolic process in Eukaryotic cells. The process involves unique trafficking event in which cytoplasmic constituents are isolated and delivered from the cytoplasm to lysosome for degradation. In addition to the basic roles in catabolism, the pathway is also utilized in the elimination of intracellular pathogen and presentation of antigen in Mammal. There are three different types of autophagy: chaperone-mediated autophagy, microautophagy, and macroautophagy. In the chaperone-mediated autophagy, degradation targets that contain specific motif are recognized and translocated into lysosomes by chaperone heat shock cognate 70 (Hsc70) and the receptor lysosome-associated membrane protein 2A (LAMP2A). In the microautophagy, lysosome directly engulfs the cytoplasmic components. In the macroautophagy, cellular constituents are isolated by the transient double-membrane–bound structure called autophagosome and delivered to lysosome by the fusion of the two membrane-bound structures. The autophagosomes originate from a small single-layer membrane structure termed Precursor Membrane Structure (PAS). Through hemi-fusions with additional membrane structures, PAS develops into a flat double-membrane structure, termed the isolation membrane, which expands and enwraps portions of the cytoplasm. As their ends meet, it results in a large double-membrane vesicle, or an autophagosome, that isolates the degradation targets. The maturation of autophagosome involves an ubiquitin-like conjugation system in Autophagy-related protein (Atg) family; in which LC3, a mammalian homologue of yeast Atg8, is conjugated to phosphatidylethanolamine (PE) by E3-like ligase Atg16(L) complex. Ethanolamine-conjugated LC3 and its orthologue GATE-16, as well as the yeast Atg8, promote tethering and hemi-fusion of membranes in vitro. Thus the conjugation system is considered essential for the autophagosome maturation. Atg16L complex is suggested to function during the elongation step, as the complex dissociates from the membrane upon the completion of autophagosome. The structure and function of those Atg proteins are extensively studied in yeast. Multiple crystal structures are determined for the fragments of the yeast Atg16 complex, which consists of Atg5-Atg12 conjugate and Atg16. The model of the overall structure has been proposed by Fujioka, et al; Atg5-Atg12 conjugate binds to N-terminal helix of Atg16, which dimerizes through its C-terminal coiled-coil domain to form a complex that contains two copies of each of the three Atg proteins. The complex associates with membranes through Atg5, and induce homotypic tethering of vesicles in vitro even in the absence of Atg8. It has confirmed in vivo and in vitro that Atg16 is required for the efficient conjugation of LC3 and PE. Moreover, the dimerization of Atg16 at the coiled-coil domain is reported to be essential for autophagy in vivo. The corresponding protein complex in Mammal consists of Atg5-Atg12 conjugate and Atg16L; thus named Atg16L complex. The mammalian Atg16L complex is similar to the yeast Atg16 complex: It associates with membranes through Atg5; dimerizes through the coiled-coil domain of Atg16L; and has E3-ligase activity toward LC3, the homologue of Atg8. However, Atg16L is significantly different from the yeast Atg16. It binds to small GTPase Rab33 at the region following the coiled-coil domain, and also has an additional WD40 domain the C-terminus. Recent reports identified a single nucleotide polymorphism in N-terminus of the WD40 domain that is associated with inflammatory Crohn's disease. Thus, those additional regions of Atg16L may be involved in the distinct mammalian-autophagic functions that are absent in yeast. The mammalian Atg16L directly interacts with Rab33b, the Golgi-resident small GTPase involved in Golgi-to-endoplasmic reticulum (ER) retrograde membrane trafficking. Rab33b interacts with Atg16L, at the region following the coiled-coil region (residues 80-200), and recruits not only Atg16L but also Atg12, which does not directly bind Rab33b. This indicates that Rab33b is able to recruit the entire Atg16L complex. Indeed, the expression of Rab33b-binding domain of Atg16L strongly inhibited autophagosome formation. However, the molecular mechanism of this recruitment and subsequent function of Atg16L complex remained unrevealed, for the structure of the mammalian Atg16L and its complex with Atg5-Atg12 is not yet known. The X-ray crystal structure was determined for the coiled-coil domain of mouse Atg16L; which had an antiparallel coiled-coil dimer in contrast to the yeast parallel Atg16. The dimerization interface had hydrophobic interactions and hydrogen bonds at the residues that were not conserved between mammal and yeast. These findings suggested that the overall structure of mammalian Atg16L complex is quite distinct form that of the yeast Atg16 complex., 総研大甲第1586号}, title = {Structural and biochemical studies on protein complexes regulating membrane traffic}, year = {} }